Properties of High Volume Fly Ash Concrete: Hsien-Hsin Hung

Properties of High Volume Fly Ash Concrete
Hsien-Hsin Hung
Thesis Submitted to the University of Sheffield
for the Degree of Doctor of Philosophy
in the Faculty of Engineering
Department of Mechanical Engineering
May 1997
"PROPERTIES OF HIGH VOLUME FLY ASH CONCRETE"
SUMMARY
This thesis presents a detailed investigation on the engineering properties and microstructural
characteristics of concrete containing a high volume of fly ash ( HVFA ). The purpose of the
project is to evaluate the concept of using relatively large volumes of fly ash in nonnal
portland cement concrete, and hence enhance the beneficial use of fly ash in ,"alue-added
products and construction.
A total of eight concrete mixtures with and ,\ithout fly ash was investigated. The
proportion of fly ash in all the HVFA concrete mixtures varied from 50 to 80 % by \yeight of
the cementitious materials, with a constant water-to-cementitious ratio of 0.40 for all the
mixtures. A high degree of workability was maintained by the use of a superplasticizer. To
optimize the pozzolanic activity in the HVFA concrete, silica fume was used in some of the
mixes. The total cementitious materials content was kept constant at 350 kg/m 3 and 450
3
kg/m respectively. The influence of the different replacement materials and two curing
regimes was studied.
The study consisted of two parts. The first part is an ex.1ensive study of the
engineering properties such as strength development, modulus of elasticity, ultrasonic pulse
velocity, swelling, and drying shrinkage at various ages up to 18 months. The depth of
carbonation of HVFA concrete under different curing regimes was also investigated. A study
of the microstructure of HVF A concretes forms the second part of the investigation. Pore
structure, air permeability and water absorption of HVF A concretes "ith different
replacement mixtures were studied. A detailed discussion dealing with the change of the
morphological phase under different curing regimes is also presented.
The results show that HVF A concretes exhibit excellent mechanical properties ",ith
good long-term strength development. Compressive strength in the range of 40 to 60 MPa
"as achieved for all the HVFA concretes at the age of 90 days. The d)namic modulus of
elasticity reached values of the order of 55 GPa at 90 days. Under similar conditions,
concretes made \\ ith both fly ash and silica fume had engineering properties which were as
good as those made with cement replaced by fly ash alone. The use of fly ash to replace both
cement and sand has the advantage of mobilizing and combining the benefits and effects of
both separate replacements. The HVFA concretes also have low permeability and exhibit
good potential characteristics to resist water penetration.
Reduction in the volume of large pores was observed with the progress of the
pozzolanic reaction. Higher HVFA concrete strength was generally associated \\ith a lower
volume of large pores in the concrete. A decrease in the levels of calcium hydroxide was seen
with progressive water curing and age in all the HVFA concretes, providing evidence of
continued pozzolanic reactivity of the fly ashes.
Various empirical relationships and design equations are presented and conclusions
are drawn at the end of each part. It is recommended that further research is required to
determine the influence on HVF A concretes of extreme curing conditions such as high or low
temperature and low moisture availability, and to improve the early strength properties of the
HVFA concretes.
II
ACKNOWLEDGEMENTS
The author would like to express heartfelt gratitude to Professor R. N. Swamy, who
gave kindly guidance and advice at all stages of the study, offered many helpful
suggestions, and comments during this investigation. Many thanks are offered to
Him and Mrs. Swamy for their thoughtful and frequent concern about the life of the
author's family during the course of this study.
Thanks are also extended to Professor J. H. Sharp, Head of Department of

Engineering materials, for permitting the use of Micromeritics Pore Size Analyser and
X-ray Diffraction facilities.
The author also wishes to acknowledge the assistance of the technical staff, in
particular Mr. N. Morton, Mr. A. Smart, of the Department of Mechanical
Engineering; and also from the Department of Civil and Structural Engineering, and
the Department of Engineering Materials.
The author is indebted to his wife, Chao-Chih, for her most encouragement,
and patient understanding throughout the research period, and the continuous
concern and constant support from his sisters and brothers.
III
Table of Contents
SUMMARy ............................................................................................................. i
ACKNOWLEDGEMENTS ..................................................................................... iii
T ABLES OF CONTENTS ..................................................................................... .iv
LIST OF TABLES .................................................................................................. ix
LIST OF FIGURES ............................................................................................... xiii
NOTATIONS AND ABBREVIATIONS ............................................................ xviii
CHAPTER ONE: IN"TRODUCTION .................................................................. 1

1-1. INTRODUCTION ........................................................................................... 1
1-2. BACKGROUND TO THE PROJECT. ............................................................. 2
1-3. AIMS AND OBJECTIVES OF THE INVESTIGATION ................................. 2
1-4. THESIS LAyOUT ........................................................................................... 4
CHAPTER TWO : LITERATURE REVIEW ..................................................... 7

2-1. INTRODUCTION ........................................................................................... 7
2-2. THE CHEMICAL COMPOSITION AND PHYSICAL CHARACTERISTICS
OF FLY ASH ................................................................................................... 9
2-3. CLASSIFICATION OF FLY ASH .................................................................. 11
2-4. HYDRATION AND POZZOLANIC REACTION .......................................... 12
2-4-1. Hydration ofpure cement compounds ....................................................... 12
2-4-2. Mechanism of hydration between fly ash and portland cement .................. 16
2-5. CHARACTERISTICS OF FLY ASH ON FRESH CONCRETE ..................... 19
2-5-1. Bleeding behaviour and setting time ......................................................... 19
2-5-2. Workability andfineness ........................................................................... 19
2-5-3. Heat of hydration ...................................................................................... 20
2-6. INFLUENCE OF FLY ASH ON HARDENED CONCRETE .......................... 21
2-6-1. Compressive strength ................................................................................ 21
2-6-2. Drying shrinkage ...................................................................................... 23
2-6-3. Creep ........................................................................................................ 23
2-6-4. Modulus of elasticity ................................................................................. 24
2-7. INFLUENCE OF FLY ASH ON DURABILITY OF CONCRETE ................. 24
2-7-1. Permeabili(v ............................................................................................. 24
IV
2-7-2. Sulphate resistance of concrete ................................................................. 26
2-7-3. Expansion due to alkali-aggregate reaction .............................................. 27
2-7-4. Carbonation .............................................................................................. 28
2-7-5. Chloride penetration ............................................ ..................................... 28
2-8. INFLUENCE OF SILICA FUME IN CONCRETE ......................................... 30
2-9. HIGH STRENGTH / HIGH PERFORMANCE CONCRETE AND HIGH
VOLUME FLY ASH CONCRETE ....................... '" ....................................... 31
CHAPTER THREE: EXPERIMENTAL PROGRAMME ............................... 38

3-1. INTRODUCTION .......................................................................................... 38
3-2. TEST PROGRAMME AND DETAILS OF CONCRETE MIXTURES ......... .38
3-3. CHARACTERISTICS OF MATERIALS ....................................................... .40
3-4. CONCRETE MIXTURE DESIGN ................................................................ .43
3-5. MANUFACTURE OF TEST SPECIMENS AND CURING .......................... .44
3-6. DOSAGE OF SUPERPLASTICIZER ( HRWR ) AND SLUMP .................... .45
3-7. TEST DETAILS ............................................................................................. 46
3-7-1. Flexural and compressive strength test...................................................... 46
3-7-2. Dynamic modulus of elasticity ...................................................... ............ -17
3-7-3. Ultrasonic pulse velocity ........................................................................... 47
3-7-4. Swelling and drying shrinkage .................................................................. 49
3-7-5. Carbonation ..... ..... '" ................................................................................. 49
CHAPTER FOUR: TEST RESULTS AND DISCUSSION OF PART 1........... 51

4-1. INTRODUCTION .......................................................................................... 51
4-2. STRENGTH DEVELOP~NT ...................................................................... 51
4-2-1. Effect of curing condition on compressive strength ................................... 52
-1-2-2. Effect of curing condition on flexural strength .......................................... 62
-1-2-3. Relationship betweenflexural strength and compressive strength. ............. 68
-1-2--1. Comparison with normal OPC concrete ............. ....................................... 7J
4-2-5. Comparison with total binder content .... ................................................... 74
4-3. COMPARISON OF STRENGTH DEVELOP~NT WITH PUBLISHED
DATA ............................................................................................................. 77
4-4. SWELLING OF HIGH VOL~ FLY ASH CONCRETE ............................ 77
4-5. DRYING SHRINKAGE OF HIGH VOLUME FLY ASH CONCRETE ......... 83
4-5-1. Effect of different mixes of drying shrinkage ...... ....................................... 83
-1-5-2. Prediction of drying shrinkage on HVFA concrete .................................... 91
4-5-3. Comparison of drying shrinkage with published data ................................ 93
4-6. DYNAMIC MODULUS OF ELASTICITY OF HVF A CONCRETE .............. 95
4-6-1. Effect of curing condition and different mixes on dynamic modulus of
elasticity .................................................................................................... 96
4-6-2. Comparison with normal OPC concrete .................................................. 101
4-6-3. Relationship between the dynamic modulus and compressive
strength .................................................................................................... 103
4-6-4. Comparison of modulus elasticity prediction ..... ....................... ............... 106
4-7 . ULTRASONIC PULSE VELOCITY OF HVF A CONCRETE ...................... 108
4-7-1. Effect of curing condition and different mixes on HVFA concrete ........... 109
4-7-2. Comparison with normal OPC concrete .................................................. 115
4-7-3. Relationship between the pulse velocity and the dynamic modulus
elasticity ................................................................................................... 116
4-8. CARBONATION OF HVFA CONCRETE ................................................... 119
4-8-1. Effect of curing condition and different replacement methods ................. 120
4-8-2 Carbonation rate coeffiCient .................................................................... 126
4-8-3. Relationship between the carbonation depth and compressive
strength ...................... " .......................................... " ................................ 128
4-8-4. Comparison with published data ...................................... ....................... 130
4-9. CONCLUSIONS ........................................................................................... 131
CHAPTER FIVE : MICROSTRUCTURE OF HYDRA TED CEMENT

PASTES .............................................................................................................. 134
5-1. INTRODUCTION ........................................................................................ 134
5-2. CLASSIFICATION OF PORE SIZE IN HYDRATED CEMENT
PASTES ........................................................................................................ 134
5-3. INFLUENCE OF POROSITY ON STRENGTH ........................................... 137
5-4. INFLUENCE OF POROSITY ON DRYING SHRINKAGE ......................... 140
5-5. INFLUENCE OF POROSITY ON PERMEABILITY ................................... 140
5-6. INFLUENCE OF POROSITY ON ABSORPTION ....................................... 142
5-7. MICROSTRUCTURE OF HARDENED CEMENT PASTE ......................... 143
5-7-1. Calcium silicate hydrate gel (C-S-H gel ) ............................................... 1-1-1
\1
5-7-2. Calcium hydroxide (CH) ...................................................... ................. 1-16
5-7-3. Calcium sulfoalumimate hydrates ........................................................... 1-17
5-7-4. Carbonates ............................................................................................. 1-18
CHAPTER SIX: EXPERIMENTAL PROGRAMME OF PART 11 ............... 149

6-1. rnTRODUCTION ........................................................................................ 149
6-2. CONCRETE MIXTURES AND CHARACTERISTICS OF MATERIAL ..... 150
6-3. TEST PARANlETERS, PREPARATION OF TEST SPECIMENS ............... 150
6-4. TEST DETAILS AND PROCEDURE .......................................................... 150
6-4-1. Mercury intrusion porosimetry study....................................................... 150
6-4-2. Air permeability ...................................................................................... 157
6-4-3. Water absorption .................................................................................... 158
6-4-4. X-r~ diffraction ..................................................................................... 159
CHAPTER SEVEN: TEST RESULTS AND DISCUSSION OF PART 11 ..... 162

7-1. rnTRODUCTION ........................................................................................ 162
7-2. PORE SIZE DISTRIBUTION OF NORMAL OPC AND HVF A
CONCRETE ................................................................................................. 163
7-2-1. Effect of different mixes and curing conditions on pore size
distribution ..... '" .................................................... ,.................................. 163
7-2-2. Maximum continuous pore diameter ofpore structure ............................. 171
7-2-3. Comparison with the total binder content on pore size distribution ......... 179
7-2-4. Total, macro-pore and meso-pore volumes.............................................. 186
7-2-5. Some relations between the pore structure and compressive strength ...... 188
7-3. PERMEABILITY OF HVFA CONCRETE ................................................... 192
7-3-1. Effect of curing condition on air permeability ......................................... 192
7-3-2. Permeability and strength development of high volume fly ash
concrete .................................................................................................... 198
7-3-3. Permeability-porosity relation in high volume fly ash concrete ............... 200
7-4. WATER ABSORPTION OF HVF A CONCRETE ........................................ 201
7--1-1. Effect of curing condition and different replacement methods ................. 202
7--1-2. Absorption-permeability relation of HVFA concrete ........... .................... 206
7-5. SOLID PHASE ANALYSIS : X-RAY POWDER DIFFRACTION ............... 208
7-5-1. X-r~ d~ffraction o.fnormal OPC and HVFA concretes with water
HI
curing ....................................................................................................... 208
7-5-2. Influence of air curing on X-ray diffraction ............................................ 217
7~6. CONCLUSION ............................................................................................. 220
CHAPTER EIGHT: CONCLUSIONS AND RECOMMENDATIONS FOR

FUTURE RESEARCH ....................................................................................... 222
8~ 1. OVERALL CONCLUSIONS ........................................................................ 222
8~2. RECOMMENDATIONS FOR FUTURE WORK .......................................... 225
REFERENCES ....•..........•................................................................................... 227
viii
List of Tables
Table 2-1 Chemical composition (% ) offly ash in various countries ...................... 10

Table 2-2. Relative proportions of hydration products in the solid portion of
cement paste ( assuming pure components) ............................................ 16
Table 2-3 Basic technical requirements of surface coatings ...................................... 30
Table 2-4. Typical high compressive strength of fly ash concrete (Unit: MPa ) ...... 32
Table 2-5 Mix proportions and engineering properties of some HVF A concretes .... 36
Table 3-1 High volume fly ash concrete test parameters .......................................... 39
Table 3-2 High volume fly ash concrete mixtures (Unit: kglm3 ) ............................ 39
Table 3-3 Chemical composition of cements and mineral admixtures ( Unit: % ) .... 41
Table 3-4 Sieve analysis of fine aggregate and coarse aggregate .............................. 42
Table 3-5 Chemical compositions offine aggregate and coarse aggregate ............... 42
3
Table 3-6 High volume fly ash concrete mix proportions ( Unit: kglm ) ............... .43
Table 3-7 High volume fly ash concrete workability ............................................... .45
Table 3-8 High volume fly ash concrete test details ................................................ .48
Table 4-1 Compressive strength ofHVF A concrete under water curing
(Unit: MPa ) ......................................................................................... 55
Table 4-2 Rate of gain of compressive strength as a % of 28 days strength
(Water curing) (Unit: % ) ................................................................... 56
Table 4-3 Compressive strength ofHVFA concrete under air curing
(Unit: MPa ) ......................................................................................... 58
Table 4-4 Rate of gain of compressive strength as a % of28 days strength
( Air curing) ( Unit: % ) ........................................................................ 59
3
Table 4-5 Influence of air curing on compressive strength-350 kglm binder
content ( Unit : % ) ................................................................................ 60
3
content (Unit: %) ................................................................................ 60
Table 4-7 Flexural strength ofHVF A concrete under water curing
( Unit: MPa ) ......................................................................................... 63
Table 4-8 Rate of gain of flexural strength as a 010 of28 day strength
(Water curing) ( Unit: % ) ................................................................... 64
Table 4-9 Flexural strength of HVF A concrete under air curing ( Unit : MPa ) ........ 65
Table 4-10 Rate of gain of flexural strength as a % of 28 days strength
IX
( Air curing) (Unit : %) ......................................................................... 66
Table 4-11 Influence of air curing on flexural strength-350 kg/m3 binder content
(Unit: % ) ............................................................................................. 67
(Unit: % ) ............................................................................................. 67
Table 4-13 Strength development of normal OPC concrete ..................................... 72
Table 4-14 Swelling strain ofHVFA concrete (Water curing)
6
(Unit: 10- mlm) .................................................................................. 79
Table 4-15 Drying shrinkage strain of HVF A concrete ( air curing)
(Unit: 10-6 mlm) ................................................................................... 84
Table 4-16 Typical % of meso pore of high volume fly ash concrete under air
curing ( Unit: % ) .................................................................................. 88
Table 4-17 Comparison between measured and predicted drying shrinkage
6
(Unit: 10- mlm) ................................................................................... 93
Table 4-18 Comparison of drying shrinkage with published data ............................. 94
Table 4-19 Dynamic modulus of elasticity of HVF A concrete ( Water curing)
(Unit: GPa) ......................................................................................... 96
Table 4-20 Dynamic modulus of elasticity ofHVFA concrete ( Air curing)
( Unit : GPa ) ......................................................................................... 96
Table 4-21 Influence of air curing on dynamic modulus of elasticity ( Unit: % ) ... 100
Table 4-22 Dynamic modulus of elasticity of normal OPC concrete ....................... 101
Table 4-23 Rate of gain of modulus as a % of28 day modulus (Water curing)
(Unit:%) ........................................................................................... 102
Table 4-24 Rate of gain of modulus as a % of 28 day modulus ( Air curing)
( Unit: % ) ........................................................................................... 102
Table 4-25 Comparison of predicting modulus of elasticity and published data under
water curing ( 28 days) ........................................................................ 107
Table 4-26 Comparison of predicting modulus elasticity and published data under
water curing ( 504 days) ...................................................................... 107
Table 4-27 Ultrasonic pulse velocity ofHVFA concrete (Water curing)
( Unit: kmlsec ) ................................................................................... 110
Table 4-28 Ultrasonic pulse velocity ofHVFA concrete (Air curing)
( Unit: kmlsec ) .................................................................................. III
Table 4-29 Influence of air curing on pulse velocity ( Unit: % ) ........................... 112
x
Table 4-30 Ultrasonic pulse velocity of normal OPC concrete ............................... 115
Table 4-31 Rate of gain of velocity as a % of28 day pulse velocity (Water curing)
( Unit: % ) ........................................................................................... 116
Table 4-32 Rate of gain of velocity as a % of28 day pulse velocity ( Air curing)
( Unit: % ) ........................................................................................... 116
Table 4-33 Carbonation depth - 350 kg/m3 mixes ( Unit: mm ) ............................. 120
3
Table 4-34 Carbonation depth - 450 kg/m mixes (Unit: mm ) ............................. 122
Table 4-35 Estimated parameters of equation C = a + KJi (t = sec ) .................. 126
Table 4-36 Comparison of carbonation depths with published data ........................ 130
Table 5-1 Classification of pore sizes in hydrated cement pastes ............................ 137
Table 6-1 A typical output ofMicromeritics pore size analyses ............................. 154
Table 7-1 Total accessible pore volume ofOPC and HVFA concrete
(Water curing) (Unit: mL/g) ............................................................ 168
Table 7-2 Total accessible pore volume of OPC and HVF A concrete
(Air curing) (Unit: mL/g ) ................................................................. 168
Table 7-3 The porosity ofOPC and HVFA concrete (Water curing) (Unit: % ). 169
Table 7-4 The porosity of OPC and HVF A concrete ( Air curing) ( Unit : % ) ..... 169
Table 7-5 The max. dv/d (log d) and threshold diameter ofOPC and HVFA
concrete (Water curing ) ...................................................................... 174
Table 7-6 The max. dv/d ( log d ) and threshold diameter of OPC and HVF A
concrete ( Air curing) ................................... '" .'" ................................ 174
Table 7-7 The macropore and mesopore ofOPC and HVFA concrete
(Water curing) (Unit: % ) ................................................................. 187
Table 7-8 The macropore and mesopore ofOPC and HVFA concrete
(Air curing) (Unit: % ) ...................................................................... 187
16 2
Table 7-9 High volume fly ash concrete permeability coefficient (Unit: 10. m ). 193
Table 7-10 Influence of air curing on permeability ( Unit: % ) .............................. 196
Table 7-11 High volume fly ash concrete water absorption ( Unit: % ) ................. 203
Table 7-12 Influence of air curing on water absorption (Unit: % ) ....... 205
Table 7-13 X-ray ditrraction ofOPC and HVFA concrete (Water curing)
( 28 days of hydration) ........................................................................ 211
Table 7-14 X-ray ditrraction ofOPC and HVFA concrete (Water curing)
( 90days of hydration ) ......................................................................... 212
Xl
Table 7-15 X-ray dim-action ofHVFA concrete (Water curing)
( 18 months of hydration) .................................................................... 213
Table 7-16 X-ray dim-action ofHVFA concrete (Air curing)
( 18 months of hydration ) .................................................................... 218
xii
List of Figures
Figure 4-1 Compressive strength ofHVFA concrete-Water curing .......................... 55

Figure 4-2 Rate of compressive strength as % of28 days strength-Water curing ..... 56
Figure 4-3 Compressive strength ofHVF A concrete-Air curing .............................. 58
Figure 4-4 Rate of compressive strength as % of28 days strength-Air curing .......... 59
Figure 4-5 Flexural strength ofHVFA concrete-Water curing ................................. 63
Figure 4-6 Rate of flexural strength as % of 28 days strength-Water curing ............. 64
Figure 4-7 Flexural strength ofHVF A concrete-Air curing ..................................... 65
Figure 4-8 Rate offlexural strength as % of28 days strength-Air curing ................. 66
Figure 4-9 Relationship between flexural strength and compressive strength ofHVF A
concrete-Water curing ............................................................................ 69
Figure 4-10 Relationship between flexural strength and compressive strength of
HVF A concrete-Air curing ................................................................... 69
HVFA concrete .................................................................................... 70
Figure 4-12 Effect of total binder content on compressive strength
(Water Curing ) ................................................................................... 75
( Air curing ) ........................................................................................ 75
Figure 4-14 Effect of total binder content on flexural strength (Water curing) ....... 76
Figure 4-15 Effect of total binder content on flexural strength (Air Curing ) ........... 76
3
Figure 4-16 Variation of swelling ofHVFA concrete-350 kg/m mixes ................... 80
3
Figure 4-17 Variation of swelling of HVF A concrete-450 kg/m mixes ................... 80
Figure 4-18 Effect of total binder content on swelling-350F/450F ........................... 81
Figure 4-19 Effect of total binder content on swelling-350SF/450SF ....................... 82
Figure 4-20 Effect of total binder content on swelling-350S/450S ........................... 82
3
Figure 4-21 Variation of drying shrinkage ofHVFA concrete-350 kg/m mixes ...... 85
3
Figure 4-22 Variation of drying shrinkage ofHVF A concrete-450 kg/m mixes ...... 85
3
Figure 4-23 Swelling I drying shrinkage characteristics-350 kg/m mixes ................. 87
3
Figure 4-24 Swelling I drying shrinkage characteristics-450 kg/m mixes ................. 87
Figure 4-25 Relationship between drying shrinkage and swelling ............................. 88
Figure 4-26 Effect of total binder content on drying shrinkage-350F/450F .............. 89
Figure 4-27 Effect of total binder content on drying shrinkage-350SF/450SF .......... 90
Figure 4-28 Effect of total binder content on drying shrinkage-350S/450S .............. 90
xiii
Figure 4-29 The curve for determining the constants a and b (350 kg/m3 mixes ) .... 92
Figure 4-30 The curve for determining the constants a and b ( 450 kg/m3 mixes ) .... 92
Figure 4-3 1 Variation of dynamic modulus of HVF A concrete ( Water curing) ...... 98
Figure 4-32 Variation of dynamic modulus ofHVFA concrete ( Air curing ) ........... 98
Figure 4-33 Effect of total binder content on dynamic modulus (Water curing) ..... 99
Figure 4-34 Effect of total binder content on dynamic modulus ( Air curing ) .......... 99
Figure 4-35 Relationship between dynamic modulus of elasticity and compressive
strength ( Water curing) ... ,.................................... " .......................... 105
Figure 4-36 Relationship between dynamic modulus of elasticity and compressive
strength ( Air curing ) ......................................................................... 105
Figure 4-37 Comparison of predicting modulus of elasticity and published data ..... 108
Figure 4-38 Variation of pulse velocity ofHVFA concrete under water curing ...... 110
Figure 4-39 Variation of pulse velocity ofHVFA concrete under air curing .......... 112
Figure 4-40 Effect of total binder content on pulse velocity ( Water curing) ......... 114
Figure 4-41 Effect of total binder content on pulse velocity ( Air curing) ............. 114
Figure 4-42 Relationship between pulse velocity and dynamic modulus of elasticity
(Water curing) ................................................................................. 117
Figure 4-43 Relationship between pulse velocity and dynamic modulus of elasticity
( Air curing ) ...................................................................................... 118
Figure 4-44 Carbonation depth of concrete - 350 kg/m3 mixes .............................. 121
3
Figure 4-45 Carbonation depth of concrete - 450 kg/m mixes .............................. 123
Figure 4-46 Effect of total binder content on carbonation depth ............................ 123
Figure 4-47 Carbonation of specimens stored in the laboratory - 180 days ............ 124
Figure 4-48 Carbonation of specimens stored in the laboratory - 540 days ............ 124
Figure 4-49 Exposure period versus carbonation depth of concrete -
350 kg/m3 mixes ................................................................................. 127
Figure 4-50 Exposure period versus carbonation depth of concrete-
450 kg/m3 mixes ................................................................................. 127
Figure 4-51 Relationship between carbonation depth and compressive strength ..... 129
Figure 5-1 Comparison of pore classifications ....................................................... 136
Figure 5-2 Relation between the compressive strength of mortar and
gel/space ratio .................................................................................... 139
Figure 5-3 Relation between compressive strength of mortar and of calculated from
the volume of pores larger than 20 run in diameter ................................ 139
Figure 5-4 The composition of the calcium silicate hydrates .................................. 145
Xl\'
Figure 6-1 Photograph of the Micromeritics Poresizer ( Model 9320 ) .................. 152
Figure 6-2 A general view of "The Leeds Permeability Cells" ................................ 157
Figure 6-3 Correction factor related to the length of the specimen ......................... 159
Figure 7-1 Cumulative pore volume curves for the 350 kg/m3 binder content OPC
and HVF A concrete ( Water curing) .................................................... 164
and HVF A concrete ( Air curing) ........................................................ 165
and HVFA concrete ( Water curing) .................................................... 166
Figure 7-5 Differential pore volume curves for the 350 kg/m3 binder content OPC
and HVFA concrete ( Water curing) .................................................... 175
and HVF A concrete ( Water curing) .................................................... 177
Figure 7-9 Cumulative and differential pore volume of 350C/450C, 350F/450F
(Water curing 28 day) ......................................................................... 180
Figure 7-10 Cumulative and differential pore volume of 350SF/450SF, 350S/450S
(Water curing 28 day) ......................................................................... 180
Figure 7-11 Cumulative and differential pore volume 350C/450C, 350F/450F
(Water curing 90 day) ......................................................................... 181
(Water curing 90 day) .......................................................................... 181
Figure 7-13 Cumulative and differential pore volume of 350F/450F
( Water curing 540 day) ........................................................................ 182
( Water curing 540 day) ........................................................................ 182
Figure 7-15 Cumulative and differential pore volume of350C/450C, 350F/450F
(Air curing 28 day) ............................................................................... 183
(Air curing 28 day) ............................................................................... 183
(Air curing 90 day) ............................................................................... 184
( Air curing 90 day) .............................................................................. 184
( Air curing 540 day) ............................................................................ 185
(Air curing 540 day) ........................................................................... 185
Figure 7-21 Relationship between compressive strength and porosity ofOPC and
HVFA concrete-Water curing ............................................................ 189
HVFA concrete-Air curing ................................................................. 189
HVF A concrete .................................................................................. 191
Figure 7-24 Relationship between compressive stren~h and meso-pore ofHVFA
concrete ................................................... " ........................................ 191
Figure 7-25 The permeability coefficient ofHVF A concrete-water curing
( 350 kg/m3 mixes) ............................................................................ 194
Figure 7-26 The permeability coefficient ofHVFA concrete-water curing
( 450 kg/m3 mixes) ............................................................................ 194
Figure 7-27 Relationship between permeability and compressive strength ofHVFA
concrete-water curing ........................................................................ 199
Figure 7-28 Relationship between permeability and compressive strength ofHVFA
concrete-air curing ............................................................................. 199
Figure 7-29 Relationship between permeability and meso-pore ofHVFA concrete
............................................................................................................................. 201
Figure 7-30 The water absorption ofHVF A concrete-water curing
( 350 kg/m3 mixes ) ............................................................................ 204
Figure 7-31 The water absorption ofHVFA concrete-water curing
( 450 kg/m3 mixes) ............................................................................ 204
Figure 7-32 Relationship between permeability and water absorption ofHVF A
concrete ............................................................................................. 207
Figure 7-33 X-ray difITactograms for OPC and HVFA concrete-Water curing
( 28 days hydration) .......................................................................... 214
Figure 7-34 X-ray diffiactograms for OPC and HVFA concrete-Water curing
( 90 days hydration) .......................................................................... 215
Figure 7-35 X-ray diffiactograms for HVF A concrete-Water curing
( 18 months hydration) ...................................................................... 216
Figure 7-36 X-ray diffractograms for HVFA concrete-Air curing
( 18 months hydration) ...................................................................... 219
Figure 8-1 Strategies of using fly ash in HVFA concrete construction ................... 223
NOTATIONS AND ABBREVIATIONS
A. Cement Chemistry Shorthand Notations and Terminology
C=CaO CC = Calcite
S = Si02
C4AH 13 = Tetracalcium Aluminate-13-hydrate
A= Al203 C3 A = Tricalcium Aluminate
F = Fe203 C4AF = Tetracalcium Aluminoferrite
-
C-S-H = C-S-H gel CS = Anhydrite
-
S = S03 Alite = Impure form ofC3S
C = C02 Belite = Impure form of C2 S
-
H=H20 C4AS H 12 = Monosulfoaluminate
-
CH = Calcium Hydroxide C6 AS 3H32 = Ettringite
-
CSH2 = Gypsum AFt = Al203-Fe203-trisulfate
C2ASH8 = Gehlenite Hydrate AFm = Al203-Fe203-monosulfate
B. Abbreviations
I L = Ignition Loss MIP = Mercury Intrusion Porosimetry

OM = Optical Microscopy SEM = Scanning Electron Microscopy
RT = Room Temperature ( ~ 25°C) OPC = Ordinary Portland Cement
T °c = Temperature in Celsius ScaleHVFA = High Volume Fly Ash Concrete
W / C = Water / Cement Ratio XRD = X - Ray Diffraction
W / B = Water / Binder Ratio .MFa = Megapascal
ACI = American Concrete Institute BSI = British Standards Institution
EPRI = Electric Power Research RlLEM = International Union for Materials
Institute Testing Laboratories
C. Pore Classification by IUPAC
Unit (Diameter) A run

Macropores = Large Capillaries > 500 >50
Mesopores = Small Capillaries 25 ~ 500 2.5 ~ 50
Micropores = Pores; Intrinsic part of C-S-H <25 < 2.5
"'111
Chapter 1 Introduction 1
CHAPTER ONE
INTRODUCTION
1-l. INTRODUCTION
Pozzolanic materials, such as fly ash ( FA), and cementitious materials such as
ground granulated blast furnace slag ( GGBFS ), are by-products of industry.
Pozzolanic materials have properties similar to those of cement, but are unable to
undergo hydration alone ; thus, they are used as partial replacement of cement, or
blended with portland cement [1,2,3,4]. In a study on synthetic fly ashes, it has been
shown that the final products of fly ash - portland cement reaction are the same as
those of portland cement alone i.e., C-S-H, C4AH13, C8AFH26 and C4ASH12 [5].
There are two major factors which influence the properties of fly ash concrete
- its intrinsic variation in chemical and mineralogical composition, and the philosophy
of mix proportioning adopted to produce the concrete incorporating fly ash [6]. In
most of the studies reported so far, the amount of cement replacement by fly ash is of
the order of 20 % to 30 %. Fly ash is the most abundant of all siliceous materials
available throughout the world. Only a very small proportion of this available fly ash
is used in concrete construction. The aim of this project is to explore the possibilities
of extending the level of cement replacement to much higher amounts so that a much
higher proportion of the available fly ash can be utilised in construction. The
investigation aims to establish the engineering and microstructure properties of this
high volume fly ash ( HVF A ) concrete, compared to that of conventional portland
cement systems. It is hoped that this increased usage of fly ash would help to reduce
disposal problems, and would also contribute, in a minor way, to a cleaner
environment.
1-2. BACKGROUND TO THE PROJECT
During the 1950's, construction of both hydroelectric and thermal power stations was
stepped up due to the sudden rise in demand for electricity. The production offly ash
as an industrial by-product from these power generating stations also gave rise to
studies connected with the use of fly ash as a cement replacement material. The
utilisation of fly ash in European countries began at about the 1960's, particularly in
the US SR and Great Britain.
The advantages of incorporating a good quality fly ash in concrete include the
improvement of the workability of the concrete, improvement of strength, water
tightness and durability at later ages, reduction in the heat of hydration, and reduction
of drying shrinkage. Probably, the foremost advantage of fly ash lies in the economy
resulting from reduction of cement content and the production of better quality
concrete. As shown in published literature, these advantages of fly ash compensate
for much of the defects of portland cement, and the use of good quality fly ash as an
essential ingredient of structural concrete is now widely accepted. From the point of
view of using relatively large volumes of fly ash in portland cement concrete, this
concept has many beneficial implications in terms of usage of waste materials,
reduction of environmental pollution, and the conservation of material resources.
1-3. AIMS AND OBJECTIVES OF THE INVESTIGATION
As mentioned before, much of the earlier research on fly ash concrete concentrated
on the development of concretes with about 20 ~ 30 % cement replacement. More
recently, in the last couple of decades, a limited amount of research has been carried
out on concretes incorporating higher amounts of fly ash, from 50 % to 60 % by
mass of portland cement [7]. However, there is the need for much more information
on such HVF A concrete to establish their properties reliably, and to give confidence
to suppliers and users. In the present study, the proportion of fly ash in the total
2
cementitious materials varied from 50 to 80 %, while water and portland cement

content were both kept at low levels.
The investigation is divided into two parts. Part I of the research deals with
the basic characteristics of concrete containing high volume fly ash. The concrete was
designed using different cementitious contents and replacement methods. In order to
investigate changes in engineering properties, the concretes were cured under two
different conditions, namely, prolonged water curing and prolonged air drying after
an initial 7 day water curing. Silica fume was also incorporated in some of the mixes,
and fly ash was used to replace both cement and fine aggregate. A superplastiscizer
was also incorporated to enhance workability, and to keep the water content low. In
addition to the strength development of high volume fly ash concrete, several other
concrete properties were also investigated. To complement the studies of engineering
properties, Part II of the project concentrates on the hydration reaction and
microstructure development of high volume fly ash concrete. Changes in pore
structure and mineralogical composition were also investigated along with
permeability, water absorption and the hydrate products in the pozzolanic reaction.
The main aim and objective of this project is to investigate the engineering
properties, microstructural characteristics and the hydrate products of concrete
containing high volume fly ash. The individual aims and objectives of each part are
given below :
Part I : Engineering properties of high volume fly ash concrete
The aims and objectives of this part include:
1. To develop mix proportioning methods for the production of high performance

fly ash concrete with a high level of replacement of portland cement with fly ash
under different curing conditions. This is intended to include both cement and
sand replacement by fly ash.
2. To study the short and long term strength development under different curing
conditions and mixture parameters of high volume fly ash concrete.
3
Chapter 1 Introduction
3. To examine the engineering properties of concrete incorporating high volumes of

fly ash by measuring changes in dynamic modulus of elasticity and ultrasonic
pulse velocity.
4. To evaluate the effectiveness of moist and dry curing regimes on expansion and
drying shrinkage behaviour ofHVF A concrete.
5. To assess the effect of pozzolanic activity on the carbonation characteristics of

concrete containing large quantities of fly ash.
6. To investigate the relationship between vanous engmeenng properties of

concrete incorporating high volume fly ash.
Part II : Microstructure mechanism and hydration reaction of high volume fly

ash concrete.
The aims and objectives Part II of the investigation are :
1. To study the relationship between microstructure and engineering properties of

the resulting of high volume fly ash concrete.
2. To study the pore size distribution and porosity in relation to the engineering
properties of high volume fly ash concrete.
3. To seek out the changes in calcium silicate hydrate ( C-S-H gel ), with curing
ages in the hydration reaction of hardened HVF A concrete.
4. To investigate the mineralogical composition of calcium hydroxide, Ca(OH)2,

crystals and the evidence of unhydrated concrete minerals such as gypsum,
C4AH 13 and ettringite in concrete incorporating high volume fly ash.
1-4. Thesis layout

The thesis is divided into eight chapters which include both parts of the study. A brief
outline of the thesis layout is given below.
Chapter 1 introduction 5
The first chapter presents a general introduction to the background and

advantages of fly ash incorporation in concrete, then describes the aims and the
objectives of each part of the present study.
The second chapter contains a literature review on properties of concrete

containing fly ash. The review begins with a brief outline of the basics of cement
chemistry and hydrate reaction of the cement paste. This is followed by a detailed
review of the influence of fly ash on engineering properties. A detailed review of high
performance, high volume fly ash concrete characteristics is also presented.
In the third chapter, the experimental programme deals with the details of test
parameters, concrete mixture design, specimen preparation and the experimental
procedure. The characteristics and properties of the raw materials used in the
experimental part of the research such as cement, fly ash, aggregate, superplasticizer
and silica fume are also reported.
In chapter four, the experimental results of tests from Part I relating to the
laboratory investigation of the engineering properties of high volume fly ash concrete
are discussed. Long and short term strength development along with the dynamic
modulus of elasticity and ultrasonic pulse velocity of different types of concrete are
investigated. Other high volume fly ash concrete engineering properties such as
swelling, and drying shrinkage are also presented. The effects of carbon dioxide from
the atmosphere on the depth of carbonation in high volume fly ash concrete is also
discussed in this section. The relationship between strength and other properties and
cementitious content for different replacement methods is also presented.
Part II of the project is covered in chapters five, six, and seven. In chapter
five, the review deals with a brief analysis of the pore structure of hydrated cement
paste. This is followed by a detailed review of the influence of porosity and pore size
distribution on the engineering properties. A detailed review of the hydrate products
of hardened cement paste is also presented.
Chapter six deals with the test details, the preparation of test specimens and
the tests conducted on the hydration and microstructure of high volume fly ash
5
,~----~
Chapter J Introduction 6
concrete. The principal tests used included mercury intrusion porosimetry, aIr
permeability, water absorption test and X-ray diffraction.
Chapter seven describes the results from the tests of Part II. This section
looks into the microstructure mechanism in order to confirm the results from the
engineering properties described in Chapter four of the thesis. This also contains a
detailed discussion describing the changes in the mineralogical phase of different
cementitious contents during the early and later periods of hydration of high volume
fly ash concrete. It also presents a discussion on the mechanisms of pore size
distribution.
Chapter eight summanzes the overall conclusions obtained from the two
experimental projects along with some recommendations for future work towards the
commercialisation of high volume fly ash concrete systems.
6
Chapfe-rLITterature Review
CHAPTER TWO
LITERATURE REVIEW
2-1. INTRODUCTION
There is a great deal of published literature on the properties of concrete containing

fly ash. It is impossible to produce here a detailed critical review of all these
publications. The aim here is to describe briefly those aspects of fly ash concrete
which are important and relevant to this study.
The use of fly ash as a pozzolanic material in combination with portland

cement for making fly ash concrete has been developed for nearly half a century. The
term " fly ash " first appeared in literature in 1937 in a paper by Davis et al ; this
paper presents results of extensive tests to determine the feasibility of using various
types of portland cement with fly ash [8]. In 1948, the use of fly ash in Hungry Horse
Dam gave an impetus, and exceptional progress has been made since in the
techniques of collecting and producing fly ash all over the world. During the 1950's,
the demand for electric power showed sudden growth, and along with this, there was
great activity in constructing both hydroelectric and thermal power stations so that
studies connected with fly ash were advanced. The standards on fly ash cement were
gradually established ( The United States ASTM Standard in 1954, Japan TIS
Standard in 1958, France NF Standard in 1959 ).
The utilisation of fly ash in European countries came at about the 1960's. In
the USSR and Great Britain, standards on fly ash were approved in 1963 and 1965
respectively. The first Australian Standard for fly ash in concrete was published in
1971. This was in the form of specified requirements of fly ash, and guidance for the
use of fly ash in concrete. The Canada Centre for Mineral and Energy Technology (
CANMET ) began work in 1985 to develop structural concrete incorporating high
7
Chapter 2 Literature Review 8
volumes of low calcium fly ash. In these investigations, an ASTM Class F fly ash was
used in proportions of 50 to 60 percent of total cementitious material [7]. Low unit
water content and a high degree of workability were obtained using high range water
reducing admixtures. Full details of the study of high volume fly ash concrete were
presented by EPRI ( Electric Power Research Institute) in 1993 [9].
Fly Ash is a by-product of the combustion of pulverized coal in thermal

power stations. It is removed by mechanical collectors or electrostatic precipitators
as a fine particulate residue from the combustion gases before they are discharged to
the atmosphere. Fly ash, besides being used as an admixture for cement and concrete,
is also used in considerable quantities as an additive in cellular concrete, artificial
lightweight aggregate, refractory materials, bricks, and in soil improvement. It is also
used in asphalt as well as in abrasives, purifiers for polluted water and fertiliser. It is
difficult to say how much fly ash is being used as an admixture for cement and
concrete in various countries of the word, but it is estimated that some 800 m tonnes
of fly ash are currently produced world-wide. In the UK alone some 15 to 20 m
tonnes offly ash are produced annually [10]. The annual consumption in recent years
is around 1.5 m tonnes in the USA, and about 1 m tonne each in Japan, France and
Germany.
The advantages of good quality fly ash are the improvement of workability of
concrete, improvement of strength, water tightness and durability of concrete at
advanced ages, alleviation of heat of hardening of concrete and reduction of drying
shrinkage, although probably, the foremost advantage of fly ash lies in the economy
resulting from the reduction of cement. When considering the fact that thermal power
stations have to incur great expense in disposing of the material as waste, the
advantages of utilising fly ash are even more marked.
8
Chapter"2 Literature Review 9
2-2. THE CHEMICAL COMPOSITION AND PHYSICAL

CHARACTERISTICS OF FLY ASH
Chemical composition of fly ash
The chemical composition of fly ash is determined by the coal used. More than 85
percent of most fly ashes comprise compounds and glasses formed from Si02,
A1203, Fe203, CaO and MgO. Kokubu [11] had tested the chemical composition of
fly ash in various countries as shown in Table 2-1, which emphasizes the many
differences in fly ash composition according to the country of production. The
average values of the important oxides are 40 ~ 58 % for Si02, 21 ~ 27 % for
A1203, 4 ~ 17 % for Fe203 and 4 ~ 6 % for CaO. These variations are largely due to
the differences in the quality of coal used in the various countries.
Physical characteristics
In general, fly ash consists of glassy spheres of sizes varying from under 1 Il m
to as large as 150 Ilm, although typical particle size distribution shows that most of
the material is under 20 Ilm [12]. The range of particle sizes in any given fly ash is
largely determined by the dust collection equipment used. Table 2-1 gives the specific
surface areas of fly ash in various countries. In this case also, a wide difference is
seen between countries, the value ranging from 1220 to 8100 cm2/g.
On the basis of optical and scanning electron micrograph observations,

several morphological categories of fly ash occur, and these may be summarized as
follows:
1. Most of the fly ash particles occur as solid glass spheres, which are mostly
transparent but may be partially devitrified.
2. Irregular masses of large size may exist either as agglomerates of small silicate
glass spheres or porous particles of incompletely burnt carbonaceous matter.
3. Magnetite and hematite occur as opaque spheres in high-iron fly ashes.
9
4. Sometimes low-calcium fly ashes may contain a small amount of hollow particles
which are either completely empty ( cenospheres ) or packed inside with smaller
spheres ( plerospheres ) [12].
Table 2-1 Chemical composition ( % ) of fly ash in various countries [11]
Country Number oj Loss Si02 Al203 Fe203 CaO MgO S03 Na20 K20 Specific
Composite on Surface
Samples Ignitiol1 (cm2 /g
Japan 12
Av. 0.73 57.96 25.86 4.31 3.98 1.58 0.34 1.49 2.15 3090
Max. 1.23 63.27 28.35 5.90 6.74 2.09 0.81 2.36 3.15 4150
Min. 0.06 53.41 22.88 2.82 1.04 1.00 0.02 0.88 1.73 1220
USA 34
Av. 7.83 44.11 20.81 17.49 4.75 1.12 1.19 0.73 1.97 3673
Max. 18.00 51.90 28.30 31.30 12.0 1.40 2.80 2.10 2.98 4795
Min. 1.00 32.70 14.60 8.50 11.1 0.70 0.30 0.22 1.28 2430
UK 14
Av. 3.86 46.16 26.99 10.44 3.06 1.96 1.59 0.90 3.26 5180
Max. 11.70 50.70 34.10 13.50 7.70 2.90 6.80 1.90 4.20 8100
Min. 0.60 41.40 23.90 6.40 1.70 1.40 0.60 0.20 1.80 2500
France 17
Av. 3.72 48.45 25.89 8.07 5.95 2.36 1.01 0.64 3.94 ---
Max. 15.15 54.05 33.40 15.30 38.7 4.45 7.00 0.85 6.00 ---
Min. 0.30 29.90 10.80 5.80 1.48 1.10 0.10 0.15 0.70 ---
Germany 9
Av. 9.65 41.13 24.39 13.93 5.06 1.85 0.77 --- --- ---
Max. 20.10 49.54 29.35 20.88 11.8 4.26 2.10 --- --- ---
Min. 1.48 34.10 21.06 8.37 2.18 0.75 0.12 --- --- ---
USSR 15
Av. --- 55.08 25.97 7.83 5.08 1.81 1.63 --- --- ---
Max. --- 62.08 37.15 12.01 10.6 2.90 3.78 --- --- ---
Min. --- 47.90 20.71 3.08 1.10 0.28 0.20 --- --- ---
10
Chapier 2 Literature Review II
2-3. CLASSIFICATION OF FLY ASH
From the standpoint of conspicuous differences in mineralogical composition and

properties, fly ash can be divided into two categories, which differ from each other
mainly in their calcium content. The first category, containing usually less than 5
percent analytical CaO is, generally, a product of combustion of anthracite and
bituminous coals. The second category, containing usually 15 to 35 percent analytical
CaO generally happens to be the product of combustion of lignite and sub-bituminous
coals. The ASTM Standard Specification for Mineral Admixtures ( C 618 ) does not
differentiate fly ashes on the basis of calcium content, although this objective is
achieved indirectly by requiring a minimum of 75 percent of the major non-calcium
oxides ( silica + alumina + iron oxide) for Class F fly ash, and 50 percent for Class C
fly ash, since the latter is higher in calcium.
The low-calcium fly ashes, due to the high proportions of silica and alumina,
consist principally of alumino silicate glass. In the furnace very large spheres of
molten glass may not get cooled rapidly and uniformly, thus crystallisation of
aluminosilicates such as silica-alumina ( Al203 • Si02 ) and mullite ( 3Al203 • 2Si02
) as slender needles, takes place in the interior of the glassy sphere. This partial
devitrification of glass in low-lime fly ashes, therefore, would account for the
presence of crystalline aluminosilicates which are relatively non-reactive. In regard to
high calcium fly ashes which may also contain significant amounts of magnesia,
alkalis and sulphates, the situation is more complex than the low-calcium ashes from
anthracite or bituminous coal. The chemical composition and reactivity of the non-
crystalline or glassy phase is different from the glassy phase typically present in low-
calcium fly ashes. Also, the principal crystalline mineral in high-calcium fly ash is
generally C3A, which is the most reactive mineral present in portland cement.
11
Chapter 2 LiteratureReview 12
2-4. HYDRATION AND POZZOLANIC REACTION
2-4-1. Hydration of pure cement compounds
Portland cement is a hydraulic cement made by heating a mixture of limestone and

clay to a temperature at which partial fusion occurs ( 1300°C '" 1500°C ). The clinker
produced is ground usually to a fineness exceeding 1500 cm2/g and mixed with 2 ~ 4
% gypsum.
The cement consists essentially of crystalline compounds ( minerals ) of

calcium combined with silica, alumina, iron oxide, and sulphate. Typically, the four
principal crystalline phases are tri-calcium silicate ( C3S, 26 % ~ 53 % ), di-calcium
silicate ( C2S, 16 % ~ 54 % ), tri-calcium aluminate ( C3A, 3 % ~ 15 % ), and tetra-
calcium alumino-ferrite ( C4AF, 8 % ~ 12 % ). The relative percentages of these
depend on the raw material used. In addition, many clinkers contain small amounts (
less than 3 % ) of free lime ( CaO ), periclase ( MgO ) and alkali sulphates. All the
compounds present in portland cement clinker are anhydrous, but when brought into
contact with water they are all attacked or decomposed, forming hydrate compounds.
The main hydration reaction eventually develops the rigid matrix of hardened cement
paste [12].
The chemistry of principal setting, hardening reaction and micromechanism

may be presented as follows :
A : Reaction involving calcium silicates
The hydration reactions between tr-icalcium silicate, di-calcium silicate and

water are as given below:
Heat of Hydration AH = - 114 kJ / mol
12
Heat of Hydration MI = - 43 kJ / mol
The principal hydration product is a colloidal calcium silicate hydrate gel ( C-

S-H gel) of colloidal dimensions. Calcium silicate hydrate occupies about 40 % to 70
% of the hardened cement paste volume. It is termed "gel" because its dimensions are
in the colloid range ( less than 100A ), and it is poorly crystalline, being essentially
amorphous. It possibly has a layer structure based on a degenerate structure of the
calcium silicate hydrate mineral tobermorite , which has a structure analogous to that
of common clay minerals, hence it has been called "tobermorite gel".
But in general the C / S ratio ofC-S-H gel might be in the range between 1.0
~ 2.0 with an average of 1.5. According to Young [13] the pore structure of a
saturated paste is more open than previously thought and that the microporosity may
only be created during drying.
The other product of hydration of the silicate minerals is about 20 to 25

percent calcium hydroxide ( CH ), which usually occurs as large hexagonal crystals,
typically 0.01 ~ 1 mm. Crystals big enough to be seen with the naked eye can
sometimes grow inside voids formed in concrete. Dehydration occurs rapidly at 400
~ 500°C. Calcium hydroxide in cement paste may vary in morphology, being found as
small equidimensional crystals ; large, thin, elongated crystals ; or variations of these
forms. Calcium hydroxide crystals are not pure but contain significant amounts of
silica as well as occluded C-S-H products, and therefore, it is better not to use values
for the pure compound. Calcium hydroxide can limit the durability of cement paste
because it is more soluble than C-S-H.
B : Reactions involving aluminate
( a ) Tricalcium aluminate
The primary initial reaction of C3A with water in the presence of a plentiful
supply of gypsum is :
13
Chapter 2 Literature Review
- -
C3 A + 3 CS H2 + 26 H ~ C3 A • 3 CS • H32 ------------- ( 2-3 )
This calcium sulfoaluminate hydrate is commonly called "ettringite". Ettringite

is a stable hydration product only while there is an ample supply of sulphate available.
If the sulphate is all consumed before the C3A has completely hydrated, then
ettringite becomes unstable and converts to another calcium sulfoaluminate hydrate
containing less sulphate.
- -
2 C3A + C3A. 3 CS • H32 + 4 H ~ 3 • C3A. CS • H12 -------- (2-4 )
This second product IS called monosulfoaluminate. The calcium

sulfoaluminates are a relatively minor constituent of a cement paste, making up 10 to
20 percent by volume. Thus, they are generally ignored due to their correspondingly
minor role in the microstructure of the hydrated cement paste. The morphology of
ettringite can be "needle-like", "wall-like", or "amorphous" structure depending on
time of curing and amount of ettringite formation. The morphology of hexagonal
hydrates can be clusters of small, plate crystals instead of well formed hexagonal
crystals.
The over reaction ofC3A can thus be represented by :
- -
C3A + CSH2 + 10 H ~ C4ASH12 ---------------------- (2-5 )
The monosulphate is a new source of sulphate ions from the solution, the
ettrignite can be formed once again :
The more gypsum there is in the system, the longer the ettringite will remain
stable.
In the absence of gypsum, the quantity of sulphate ions present in cement is

insufficient, which can lead to flash set. As a consequence, the reaction takes place
between C3A and the water as following:
l-l
2C3A + 21 H ~ C4 AH 13 + C2 AH8 ----------------------- ( 2-7 )
These calcium aluminate hydrates ( C2AH8 and C4AH13 ), which are

members of the AFm group and closely related structurally to monsulfate, are
metastable phase and change gradually to another more stable cubic hydrate C3AH6 ;
C3AH6 is a member of the hydrogarnet solid solution series.
C4AH 13 + C2AH8 ~ 2C3AH6 + 9 H --------------------- (2-8 )
(b ) Tetra - calcium alumino-ferrites (Ferrite phase) --- In this case, the AIF
ratio is not the same as that of the parent compound.
The hydration of tetra-calcium aluminoferrite is not as well comprehended but

forms the same sequence of hydration products as does C3A.
In the presence of gypsum, the reactions are slower and evolve less heat. The
hydration between C4AF and sulphate ions may be represented as :
- -
3 C4AF + 12 CSH2 + 110 H ~ 4 [ C6 (A,F) S 3H32 ] + 2 (A,F ) H3 ----- (2-9 )
- -
3 C4AF + 2 [C6 (A,F) S 3H32] + 14 H ~ 6 [C4 (A,F) SH12] +2 (A,F) H3
----- ( 2-10 )
In the absence of sulphate ions, sulphate-free AFm phase will be formed

analogous to Eq 2-7. The thermodynamically stable phase is C4( A,F )H6,
C4AF + 16H ~ 2 C2 ( A,F ) H8 ----------------------- ( 2-11 )
C4AF + 16 H ~ C4 (A,F) H13 + (A,F) H3 ------ (2-12)
The above equation such as ( A,F ) means that A1203 and Fe203 occur
interchangeable in the compound.
15
Chapter 2 literature Review 16
The volume of the hydration products in a full hydrated and saturated paste is
listed in Table 2-2.
Table 2-2. Relative proportions of hydration products in the solid portion of

cement paste ( assuming pure components) [13]
Compound Hydration products Volume in Volume in

pure paste ** cement paste ***
C3 S C-S-H gel 66% 32%
CH 34% 16%
C2 S C-S-H gel 87% 20%
CH 13% 3%
C3 A C~ACSH12 * 100% 19%
C4 AF # 100% 8%
Minor oxides Hydroxides --- 2%
(asMgO )
Remarks:
1. * Assumed stable phase in mature pastes.
2. ** Expressed as percentage of solid products. A density of2.25 g fcm 3 was

assumed for both CH and C-S-H gel.
3. *** Assuming a typical composition Type I for portland cement.
4. # Hydration ofC4AF assumed analogous to C3A.
2-4-2. Mechanism of hydration between fly ash and portland cement
Fly ash is a pozzolanic material. According to ASTM C-595, a pozzolan is defined

as "a siliceous or siliceous and aluminous material which in itself possesses little or
no cementitious value but which will, in finely divided form and in the presence of
moisture, chemically react with calcium hydroxide at ordinary temperature to form
compounds possessing cementitious properties".
Pozzolans are rich in Si02 and poor in CaO, whereas the opposite holds for
clinker and portland cement. Calcium hydroxide is produced from the excess CaO
16
during the hydration of cement constituents, and serves as the mam source of
alkalinity for the pozzolanic activity, i.e., for the conversion of the Si02 of the
pozzolan into C-S-H. The pozzolanic reactions of various pozzolans have been
examined in several investigations previously [11,14,15,16]. Guillaume [1 7] found a
correlation between the decrease in calcium hydroxide, produced by hydration of
cement, and the extent of pozzolanic reaction, which suggested that the pozzolanic
reaction of fly ash starts at about 14 days. Bache et a1 [18] stated further that when
fly ash absorbs calcium hydroxide, the absorbed calcium hydroxide appears to form a
membrane over the fly ash particles and a film of water exists at the interface between
the membrane and fly ash particles.
A study on synthetic fly ashes suggested that the initial reaction products
appeared to be poorly-crystallized hydrates of calcium silicate or calcium aluminate.
After one week of curing, a tetracalcium aluminate hydrate ( C4AHx ) was identified
along with some fine needle-like crystals which were probably ettringite. After eight
weeks, C-S-H was identified. The final products of fly ash reaction are the same
those of portland cement. So, in very simplified terms, pozzolanic activity can be
descried by the following reactions [5][19] :
C+H ~ C H ------------------------------------------- ( 2-13 )
2 S+3CH ~ C3S2 H3 ------------------------------ ( 2-14 )
A + F + 8 CH + 18 H ~ C8AFH26 ---------------- (2-15 )
A + CSH2 + 3 CH + 7 H ~ C4ASH12 ------------- (2-16)
A + 4 CH + 9 H ~ C4AH 13 -------------------------- ( 2-17 )
It is worth noticing that the reaction of fly ash in portland cement coincides
with the reaction of portland cement. The main difference between the portland
cement hydration reaction and the pozzolanic reaction lies in the former, where the
entire quantity of each basic oxide of portland cement, i. e., of C, S, A and F, reacts
in the form of the compounds, C3S, C2 S, C3A, and C4 AF , in which they occur in the
clinker, whereas in the latter, only the amorphous forms of these pozzolanic oxides
17
may react with lime solution ( CH ). Another difference is that the pozzolanic
reactions of fly ash may be initiated by the adsorption of calcium hydroxide, produced
by cement hydration, on the surfaces of the fly ash particles. The adsorbed calcium
hydroxide appears to form a membrane over the fly ash particles and a film of water
exists at the interface between the membrane and fly ash particles [20].
It should be recognised that, because of the compositional heterogeneity of fly

ashes and cements, and the heterogeneous nature of their pozzolanic reactions,
reactivity is not a unique property of a fly ash. Even with a single cement, standard
curing conditions and mix proportions, reactivity will probably vary with the progress
of the reaction. Further, reactivity does depend on the intrinsic and extrinsic factors
[6,21,22,23]. Intrinsic factors include the physical and chemical properties of the fly
ash. Extrinsic factors include the properties of cement, curing conditions and
environment, mix proportions, cement / fly ash ratio and water / cementitious ratio
etc. In most studies, however, the relationships between the inorganic constituents of
fly ash reactivity and hydrate reaction are still not well-defmed, as will be apparent
from this review.
18
2-5. CHARACTERISTICS OF FLY ASH ON FRESH CONCRETE
2-5-1. Bleeding behaviour and setting time
Hydration occurs right after the cementitious materials come into contact with water.
The hydraulic cementitious reaction progresses to make the cementitious system
harden and develop strength. Basically, setting behaviour is the result of cementitious
grains dispersing and hydrating with water, and gradually forming a solid / liquid
suspension of various size hydrates. As cement hydrates, the inner structure of the
cementitious paste is further reinforced, and becomes a network structure which
makes the cementitious material set and gain strength.
Bleeding is mainly due to the amount of mix water being greater than that
required for complete cement hydration. The phenomenon is caused by the solid /
liquid suspension being generated after the mixing of cement and water. Except for
the adsorption of water due to the electrical double layer and diffusive water layer,
the extra water escapes from the structural network of cementitious pastes, which
explains the mechanism of bleeding. Concrete using fly ash is generally reported to
show reduced segregation and bleeding, and to be more satisfactory when placed by
pumping than plain concrete placed under the same circumstances.
2-5-2. Workability and fineness
The pozzolanic activity of fly ash has a close relationship with its fineness. The small
size and the essentially spherical form of the particles comprising fly ash usually
influence the rheological properties of cementitious pastes, causing a reduction in the
amount of water required for a given degree of workability from that required for an
equivalent paste without fly ash [24]. For a constant workability, the reduction in the
water demand of concrete due to fly ash is usually between 5 and 15 % by
19
Chapler 2 Literature Review 20
comparison with a portland cement only mix having the same cementitious material
content. There are many reports confirming the workability and fineness
characteristics offly ash. As early as 1937, Davis et al made this clear in their paper
[8]. Kokubu and Yamada [19] showed that unit water-content ratio was lowered
with increasing fineness and that the unit water-content ratio exceeds 100 percent
when specific surface area was lower than 2000 cm2/g with almost no effect on
workability. Meanwhile, they cited the results from Homms et aI, who collected 14
varieties of fly ash varying in fineness between 1000 ~ 4150 cm2/g from a thermal
power plant and performed tests on mortar holding cement-fly ash quantities
constant, replacing 25 percent of the cement with fly ash and determining unit water
contents to obtain equal consistency. Cabrera and co-workers [25] have reported on
measuring the effects of seven fly ash on concrete workability by using Tattersall
Two Point Test. The test programme shows that fly ashes vary in their ability to
improve concrete workability. There was no question that the fineness values of a fly
ash affected the workability of the concrete. However, the standard value of fineness
is significant in providing a criterion in planning fly ash collecting facilities. Table 2-1
also gives examples of the specific surface areas of fly ash in various countries.
2-5-3. Heat of hydration
Portland cement is in a high energy state because it is produced from clinkers in a

burning process in a kiln at a high temperature. When it comes into contact with
water, the clinker converts to a low energy state, and heat escapes during the
hydration process. The heat of hydration of cement may be as high as 180 J / kg so
that in structural concretes there might occur a temperature rise of 135°C if no heat
were lost. Fortunately, in most concrete structures heat is lost to the environment
almost as fast as it is generated so that no problems of thermal stresses result. In mass
concrete, however, portions of the interior may remain in an essentially adiabatic
state for a period of weeks after the concrete is placed [24]. The pozzolanic reaction
of the aluminosilicates in fly ash with the calcium hydroxide liberated by the C 3 S and
20
C2 S components of portland cement takes place more slowly than C3S reaction and
the approximate reaction rate ofC3S, This combined with the dilution of the portland
cement component by fly ash results in a reduced rate of heat evolution, and reduced
ultimate heat of hydration. Sturrup et al [26] used a conduction calorimeter, and
found that fly ash could be used successfully in lieu of both ASTM Type II and IV
moderate and low-heat cements to control temperature rise and thermally induced
cracking in mass concrete. The use of fly ash to reduce the heat of hydration of
concrete is well established. The US Bureau of Reclamation ( USBR ) has conducted
numerous experiments concerning heat of hardening of concrete containing fly ash.
Malhotra et al also cited the results from Eifert who showed the effect of fly ash and
a calcined diatomaceous shale on the temperature rise of mass concrete, and found
that for a given total cementitious materials content, high volume fly ash concrete
exhibits a substantially lower temperature rise than conventional concrete [24,27].
2-6. INFLUENCE OF FLY ASH ON HARDENED CONCRETE
2-6-1. Compressive strength
The factors affecting the development of concrete cementitious strength include its
internal water content, curing condition, mix proportioning and admixtures. The
particles of fly ash in a dry state often present a floc structure from the mutual
attraction of the particles ; when concrete is mixed using fly ash which has a floc
structure, or present as agglomerates, the particles of fly ash are flocculated and are
difficult to disperse. Meanwhile, the contribution which fly ash makes to strength
development of concrete is the result of pozzolanic reaction between the soluble
silica and alumina from the glass phase of the fly ash, and lime, to form calcium
silicate and calcium aluminate hydrates. Moreover, in addition to the effect of
chemical reaction, fly ash has a physical effect of improving the microstructure of the
hydrated cement paste. The main physical action is that of packing of the fly ash
21
particles at the interface of coarse aggregate particles. Much research has shown that
proper percentage replacement of portland cement in concrete results in comparable
strength between normal OPC and fly ash concrete [12,14,21,24,28]. In
consideration of early strength fly ash concrete, it has been reported that concrete
mixes with normal weight and lightweight aggregates containing 30 percent by
weight of cement replacement of fly ash can be designed to have adequate early one-
day strength for structure applications. Swamy et al [21] and Cripwell et al [29] have
reported on compressive strengths when a part of normal portland cement is replaced
with fly ash ; these studies show that the later-age strength development of fly ash
concrete is higher than that of normal OPC concrete resulting in an 18 percent
increase in compressive strength at 300 days. Thomas and Matthews [30] also have
reported on investigating existing concrete structures built with fly ash concrete. The
age of structures ranged from 2 to 33 years. The results indicated that in all cases the
fly ash concrete has greater long term strength by comparison with the OPC concrete
in the same structure. These effects of sand replacement on early age and long term
fly ash concrete strength also suggests that the superior performance of these
mixtures compared to normal OPC concrete mixtures is attributable to the
densification of the paste structure due to pozzolanic action between the fly ash and
calcium hydroxide liberated as a result of hydration of cement [31].
In recent years there has been considerable interest in the production and use
of very high strength concrete ( 50 ~ 60 MPa ) at 28 days. This type of concrete is a
particular example of the combined practical use of fly ash and water-reducing
admixtures [24]. An extensive scientific and engineering study ( 1990 - 1993 ) by
EPRI ( Electric Power Research Institute) has demonstrated the use of high volume
fly ash ( HVF A ) concrete produced for structural applications. High volume fly ash
concrete has adequate early age strength, and very good later age strength ; and, the
ratio of the flexural to compressive strengths, and of splitting-tensile to compressive
strength are comparable to the values for conventional portland cement concrete [9].
22
2-6-2. Drying shrinkage
The volume of a cementitious paste varies with its water content. Drying will cause a
volume decrease, i. e., drying shrinkage ( or shrinkage ), and wetting causes volume
increase, i. e., swelling. Consequently, the rate of removal of water from the
cementitious paste controls the rate of shrinkage of the cementitious paste. All
factors which affect drying, such as temperature, humidity, pressure and arr
movement, will also influence shrinkage. Also, the factors that affect the physical and
chemical properties of the paste, such as cementitious composition, admixtures,
curing temperature, and water / cementitious ratio, will also affect shrinkage.
2-6-3. Creep
Some studies have shown that fly ash increases the creep of concrete [28]. However,
these investigations involved direct replacement of fly ash for cement, and produced
lower strength at loading. Since creep is influenced by compressive strength and the
modulus of elasticity of the concrete, higher creep strains were found in fly ash
concrete at early age loading when strength was low, however the rate of creep
decreased at later ages.
When similar 28 day strength is compared, creep of fly ash concrete is lower
than its plain concrete equivalent. Investigations by Bamforth [32] on mass concrete
showed that a reduction of 50 % in creep can be obtained when cement is replaced by
about 30 % fly ash. Ross [33] had obtained 15 % reduction with 25 % fly ash
replacement.
The high volume fly ash concrete appears to produce low creep strains
compared to those of normal concretes. The low creep strains could be the result of
a large portion of fly ash remaining unreacted in the concrete, and thus acting as fine
aggregate, providing increased restraint against creep. The long-term creep behaviour
of concrete containing high volume fly ash and superplasticizer has been described by
23
Cripwell et al [29]. The combined effect of high volume fly ash and superplasticizer is
to decrease the creep of concrete stored in water or stored in air; after 300 days of
loading, creep was 50 % less than that in the case of portland cement concrete.
Similar, long-term creep results by Alexander et al [34], showed that the creep of
plain concrete specimens loaded at stress / strength ratio of 0.25 and 0.4 were found
to be 50 % greater than that of fly ash concrete at 2 years, and the situation remained
unchanged at 6 years.
2-6-4. Modulus of elasticity
Modulus of elasticity is very important from a design point of view in reinforced and
prestressed concrete, since it can control concrete behaviour as much as compressive
strength, particularly in structural elements subjected to flexure. The modulus of
elasticity of concrete follows a pattern similar to compressive strength. Fly ash
concrete is known to exhibit similar or slightly higher modulus of elasticity than
portland cement concrete [12,24,26,35,36]. Lorig term, higher strength concrete can
help to reduce deflections under load, and these can be anticipated when fly ash is
used in concrete. A report by EPRI [9] on high volume fly ash concrete claims that,
the generally high modulus of elasticity values of this concrete seems to be due to the
densifying effect of the unhydrated fly ash particles acting as fine aggregate.
2-7. INFLUENCE OF FLY ASH ON DURABILITY OF

CONCRETE
2-7-1. Permeability
Factors affecting permeability amongst others, include the presence of large voids in
the hydrated cement, microcracks at the aggregate-cementitious paste transition
24
zone, water / cementitious ratio and curing efficiency. Unless curing conditions are
especially poor, the impermeability of concrete containing fly ash is improved. The
pozzolanic reaction of fly ash in concrete, which produces calcium silicate hydrate,
tends to fill these unoccupied spaces to form a product with decreased permeability,
which also results in improved corrosion resistance.
A number of investigations has been made on the influence of fly ash on

relative permeability. Manmohan and Mehta [37] showed that the strength
improvement as a result of pozzolanic reaction is accompanied by, and partly results
from, a reduction in the volume of large pores, and found that in cements containing
fly ash, the pore-size refinement and a drastic drop in permeability of the cement
paste ( from 11 ~ 13 x 10- 11 to 1 x 10- 11 m/sec ) occurred during the 28 to 90
days curing period. Kokubu [11] also cited results on concrete permeability which
agrees with experimental data from Davis who compared permeability coefficients of
concretes of unit cement contents of 225 kg/m3 at the age of 6 months and stated
that the coefficient of concrete with 30 % of cement replaced with fly ash was 1 / 5
of that of concrete without fly ash, while with 50 % replacement, the coefficient was
reduced to 1 / 12. This is also consistent with the results of Davis, cited by Berry and
Malhotra on permeability of a concrete pipe containing fly ash, which was higher than
that of plain concrete at 28 days, and was substantially lower after 6 months of curing
[24]. Subsequently, Malhotra [14] further added that the high performance, high
volume fly ash concrete tests performed under uniaxial flow conditions under a
-13
pressure of 27 MPa indicated the permeability to be of the order of 10m / sec. It is
clear from these investigations that the permeability of the concrete is directly
affected by hydration reaction, and that the low permeability of the concrete is due to
the low porosity of the portland cement / fly ash system with gel pores being
discontinuous.
25
2-7-2. Sulphate resistance of concrete
Most commonly, sulphate attack is a phenomenon in which the sulphates from the
environment surrounding the concrete react with reactive calcium aluminate phases
within the hardened concrete, resulting in internal expansion due to the formation of
the crystalline compound ettringite, and yields a greater volume causing
deterioration. The replacement of a portion of the portland cement with fly ash on the
sulphate resistance of concrete can be conveniently explained in terms of two
mechanisms. First, the pozzolanic reaction reduces the amount of calcium sulphate,
which consequently decreases the amount of gypsum formed. The second
mechanism, the concrete with fly ash in the mixture can decrease the proportion of
the larger pores which results in a less porous matrix, and decreases the total amount
oftricalcium aluminate (C3A) content of the cement.
In recent years, several studies have tried to document the effect of fly ash on
the sulphate resistance of concrete [35]. Mehta [38] and Carrasquillo et al [35]
indicate that the effect of fly ash on the sulphate resistance of concrete is related to
the composition of both the crystalline and amorphous phases. This phase dependent
interaction can be used to expand on the R-value concept to reliably predict the effect
of fly ash on the sulphate resistance of concrete. USBR ( United States Bureau of
Reclamation) guidelines include a resistance factor (R-value ) defined as :
CaO%-5
R= ---------- ( 2 - 18 )
Fe203%
and found that using cement replacement levels of 25 % fly ash with R-values less
than 1.5 was found to improve the sulphate resistance of the concrete, and suggested
that with the calcium oxide ( CaO ) content less than 5 % and low ferric oxide (
Fe203 ) contents in the concrete, sulphate attack is reduced. Tikalsky and
Carrasquillo [35,39] have given the same conclusion. In 1993, EPRI [9] reported the
results of sulphate resistance by using 12 different power plant fly ashes on high
volume fly ash concrete studies. From this work it was concluded that all of the fly
ashes tested greatly improved sulphate resistance.
26
2-7-3. Expansion due to alkali-aggregate reaction
This undesirable expansion of concrete is due to reaction between the aggregate and
hydroxyl and alkali ions in the pore solution forming a gelatinous alkali silicate, which
is capable of absorbing large quantities of water. The consequent expansion disrupts
the concrete and reduces both the strength and elasticity properties of concrete and
the structural integrity of concrete structures [40,41,42]. Fly ash with siliceous
constituents may react with alkalis in the cement to form products which take up
water and swell, and can minimize or prevent the expansion [28]. The use of
pozzolans to prevent alkali-aggregate reaction has been summarized from the results
of Pepper and Mather by Plowman and Harrogate. They concluded that the
effectiveness of different fly ashes in reducing long-term expansion varied widely, and
suggested that the effectiveness may be dependent upon their alkali content or
fineness [40].
Research at Canada Centre for Mineral and Energy Technology (CANMET )

[9,43,44] has shown that the use of high volume fly ash concrete is highly effective in
(
inhibiting the expansion due to alkali-aggregate reaction. The rationale for reduction
in the expansion due to alkali-aggregate reaction is the extremely low permeability of
the high-volume systems, and the reduced pH of the pore solution. Similar results by
Alasali & Malhotra [45] showed that the expansion of the concrete incorporating 58
% fly ash, after 275 days of exposure under the various test conditions, is
insignificant.
While it is clear that some fly ashes are effective in controlling alkali-
aggregate expansion, it is questionable whether the strength losses caused by
replacement of the cement by fly ash would be tolerable for more than a limited
number of applications. The use of such large amounts of fly ash to reduce alkali-
aggregate interaction, if required, would certainly demand reproportioning of
concrete if acceptable early strength were to be attained.
27
2-7-4. Carbonation
Carbonation is the process whereby carbon dioxide from the atmosphere penetrates
the concrete, converting the hydroxides to carbonates and lowering the alkalinity of
the protective concrete surrounding the steel, leading to dissolution of the protective
oxide layer on the steel bars and to reinforcement corrosion [5,45]. There are a large
number of factors that affect the rate and extent of carbonation, such as, the
concentration of Ca(OH)2 produced by the hydration of portland cement as well as
C-S-H, which contains CaO. With concrete that contains portland cement and fly ash,
there is a further increase in the number of carbonating variables when some of the
cement is replaced by fly ash [46]. Browne [47] cited the results from Mattews,
obtained at the Building Research Establishment, who showed that after 10 years, fly
ash concrete carbonates approximately to the same extent as normal OPC concrete,
with the rate of carbonation found to be related to the strength of the concrete.
Langleyet al [27] carried out tests on large concrete blocks containing high volumes
of low-calcium fly ash and showed that, for concretes of similar 730 days strength, fly
ash concrete carbonates to a lesser depth than the OPC concrete. After 7.5 years
cores of high volume fly ash concrete examined showed average carbonation depth
less than 8 mm [44].
Contradicting results have been reported by Papadakis et al [5] and

Kawamura & Haque [48] ; they indicated higher rates of carbonation in fly ash
concrete than that of the control normal OPC concrete. As both concretes had the
same value of water-to-binder ratio, a pozzolanic cement concrete had significantly
higher carbonation depth due to its greater porosity, and also due to the smaller
concentration of carbonatable constituents.
2-7-5. Chloride penetration
The main cause of reinforcement corrosion is the loss of efficiency of the concrete
cover, which protects the steel and keeps it in a passive state. Chloride ions have the
28
special ability to destroy the passive oxide film of steel. The chloride ions which
penetrate into the concrete arise either from internal mix constituents or external
sources. Internal mix constituents include CaCl2 added as an accelerating admixture
and other admixture compositions. External sources are contact with sea water, de-
icing salts and structures exposed to the splash zone and marine environment.
The paths that allow chloride ions to intrude depend upon the pore structure
of the concrete. The intrusion process of chloride ions occurs by external hydraulic
pressure that bring the ions through the pores into concrete. The amount of chloride
required to initiate corrosion depends on the pH of the solution in contact with the
steel, and comparatively small quantities are needed to offset the basicity of portland
cement [13]. Even at high alkalinities, the intrusion of chloride will be accompanied
by a chemical reaction with the tricalcium aluminate to form calcium-chloroaluminate
hydrate. There are several strategies that have been developed to prevent corrosion
of reinforcing steel, based largely on studies carried out on bridge-decks. The
possible strategies can be divided into several categories such as reduction of
permeability of concrete, protective coatings on the concrete, protective coatings on
the steel, and suppression of the electrochemical process. Swamy and Tanikawa [49]
have given recommendations for the selection criteria of surface coatings, as shown
in Table 2-3. The effect of using fly ash in concrete is to reduce the amount of
chloride intrusion which can arise due to a lower porosity and high tortuosity of the
pore channels inhibiting ionic diffusion. Dhir et al [50] found that it is not the fly ash
quality that effects chloride diffusion but the quantity of fly ash in the concrete.
Haque, Kayyali, and Gopalan [51] also observed that the presence of fly ash
significantly reduced the concentration of free CL- ions in concrete when the chloride
ingress is from the surrounding sodium chloride solution. The rapid chloride-ion
diffusion tests performed on HVFA concrete indicated very low chloride-ion
diffusion at 1 year, ranging from 119 to 179 coulombs confirming the excellent
resistance ofHVFA concrete to chloride-ion penetration [9].
29
Table 2-3 Basic technical requirements of surface coatings [49]
Characteristics of properties Descriptions

Chemical resistance IeResistance to salt ( salt spray, dry / wet sal1
spraycycle, immersion in sea water ).
~ Acid resistance.
Weathering resistance ~ Outdoor exposure ( ultra violet light, ozom
variable temp / humidity).
~ Heat resistance.
• Resistance to water permeability.
Resistance to expansive forcel- Freezing and thawing.
• Alkali silica reaction.
Engineering properties ~ Crack bridging ability.
~ Adequate elastic / fatigue resistance ( to reSIS
crack opening / closing ).
• Strain capacity - stress - strain behaviour.
~ Adhesion strength ( film continuity).
• Abrasion Resistance.
Ie Thermal Compatibility.
Diffusion resistance Ie Water, air ( oxygen, carbon dioxide ), chloridt
ions, water vapour ).
• Air diffusion at high pressure.
Economic requirements ~ Ease of application.
~ Long service life.
• Pleasant appearance.
2-8. INFLUENCE OF SILICA FUME IN CONCRETE
Silica fume is the by-product of electric arc furnaces during the production of silicon
metal or ferrosilicon alloys. Ferrosilicon alloys containing 50 ~ 70 % Si are the most
. important alloys of silicon. The raw materials used for the production of ferrosilicon
are usually 3 '" 6 inch pieces of crushed quartz, coke, and iron ore pure silicon metal
is produced at temperatures of the order of 2000°C, in the areas between electrodes
submerged in the charge. About 10 '" 15 % quartz in the charge is lost in the form of
Si and SiO vapours, which react with air to form Si02· The fine particles are
collected by filtering gases escaping from the arc furnace. The chemical compositions
of the condensed silica fume from the ferrosilicon industry corresponds, generally, to
85 '" 92 % Si02, 0.5 '" 3 % Fe203, 1 ~ 3 % alkalis, 1 ~ 2 % carbon, and minor
30
amounts of Al203, CaO, and MgO [52]. These percentage compositions depend on
what is being produced in the furnace. The particle size analyses show individual
spherical particle shape with a number of primary agglomerates, and amorphous
particles range from 0.01 to 0.3 ~m, with about 70 % < 0.1 ~m. Nitrogen
adsorption measurements show 15 - 20 m2/g surface area of the material.
Areas of application where silica fume has shown particular advantage, and
some of the results from these investigations have been recently published in a
comprehensive review [15,43,53,54,55,56]. There is strong evidence that the use of
silica fume increases the homogeneity and decreases the number of large pores in
cement paste, and results in a denser interface between cement paste and coarse
aggregate. These studies have demonstrated that the pozzolanic activity effect of
silica fume in concrete is faster than ordinary portland cement, because of its high
surface area, and the high amount content of amorphous silica in silica fume. Many
authors foresee a wide-ranging application of silica fume if it is used as a strength
accelerator in portland cement concrete containing fly ash [52].
2-9. HIGH STRENGTH I HIGH PERFORMANCE CONCRETE

AND HIGH VOLUME FLY ASH CONCRETE
The factors affecting the development of cement paste strength include its internal
water content, voids in the matrix, and interface organisation. Reducing water
content and voids as well as compressing the interface may produce high strength /
high performance concrete. Concrete is commonly used in constructing building
frames. Recent developments in high strength / high performance concrete have
extended its use to tall buildings [57].
The name "high strength / high performance concrete" implies that, in many
cases, it is high durability that is the required property, although, in others, it is high
strength, either very early, or at 28 days, or even later. In some applications, a high
modulus of elasticity is the property sought.
31
Although both properties of high strength and low permeability are not
necessarily concomitant, high strength requires a low volume of pores, especially of
the large capillary pores. The way to have a low volume of pores is, in general,
achieved by the use of mineral admixtures, such as fly ash, slag or silica fume, which
fills the space between the cement particles and between the aggregate and the
cement particles.
Indeed, high strength / high performance concrete can be said to be a logical

development of concrete containing different mineral admixtures which change the
internal structure of concrete. While there exists no standard, or even typical, mix
proportion of high strength / high performance concrete, it is useful to present
information on several successful mixes, and this is shown in Table 2-4 [58]. Several
of these mixes contain, in addition to portland cement and fly ash, other cementitious
materials. There is an economic advantage is using these various cementitious
materials, partly because they are more effective than portland cement, but also
because they allow a reduction in the quantity of cement.
\
(
Table 2-4. Typical high compressive strength of fly ash concrete [58J
(Unit: MPa)
Mixture Type Age (days) Wet 7 day wet / Air

kg/m3 aIr
350 ope 28 55.7 57.9 52.9
90 60.8 63.2 59.6
260 68.5 71.5 62.2
300 ope + 20 SF + 30 FA 28 60.1 63.6 60.1
90 75.5 77.3 71.1
260 78.3 80.1 73.0
250 ope + 20 SF + 80 FA 28 59.4 64.0 57.1
90 67.8 72.7 61.0
260 71.4 76.2 63.0
250 ope + 20 SF + 80 slag 28 67.9 70.9 61.7
90 74.8 78.4 67.5
260 84.4 88.0 70.1
32
The importance of fly ash cannot be exaggerated ; it is no longer a cheap

substitute for cement, nor an extension or an addition to the mix. The variability of
the properties of fly ash was mentioned in the preceeding section. The effects of fly
ash are activated by the pozzolanic reaction when portland cement is used with it.
Moreover, in addition to the effect of chemical reactions, fly ash has a physical effect
of improving the microstructure of the hydrated cement paste.
Concrete containing high volume low-calcium fly ash ( ASTM Class F ) has
been used in various parts of the world during the last 10 years. The first use of this
type of HVF A concrete is believed to have been in roller compacted dams and
highway base courses. Roller compacted concrete has very low unit water content,
and lacks the workability required for placement as structural concrete.
It was mentioned before that structural concrete incorporating high volumes

of low-calcium fly ash was developed at CANMET in 1985 [7]. The initial research
focused on developing a HVF A concrete mixture suitable for ready-mixed operation,
having a relativity low temperature rise, for use in massive unreinforced structural
elements. Extensive testing has shown that the above type of HVF A concrete made
with ASTM Type II cement has high unit weight, satisfactory early-age compressive
strength, high modulus of elasticity, and high later-age strength. In this type ofHVF A
concrete, cement content is kept at about 150 kg/m3 and the ratio of fly ash to total
cementitious material content is increased to about 55 %. As the water content of
these mixtures is less than 120 kg/m3, the workability is achieved by using large
dosages of superplasticizers.
Swamy and Mahmud [59] also introduced such HVFA concrete at the
international symposium on the Fly Ash, Slag, Silica Fume and Natural Pozzolans in
Concrete of 1986. Subsequently, W. S. Langley and Association limited, Halifax,
Nova Sotia undertook in a large HVFA concrete block, 3.05x3.05x3.05 m, to
monitor the long-term strength development of HVF A concrete. The water-to-
cementitious material ratio of these HVF A concrete ranged from 0.27 to 0.49, and
3
their total cementitious material content ranged from 400 to 225 kg/m . Test results
were presented in 1989 and 1992 respectively [27,60]. The test data on strength
33
Chapter 2 Literature Review 3-1
properties, modulus of elasticity, drying shrinkage, creep, and freeze-thaw durability

indicate that concrete incorporating low cement content and high volume of fly ash
compares favourably to normal portland cement concrete. Also, limited tests
performed to date indicate that the permeability of concrete incorporating high
volumes of fly ash is also very low.
An investigation undertaken on HVFA concrete made with a combination of

fly ash and silica fume was reported by Giaccio and Malhotra [61]. From the test
results, it was concluded that HVFA concrete has excellent mechanical properties
and satisfactory resistance to repeated cycles of freezing and thawing.
Xu and Sarkar [62] investigated the microstructure of HVF A concrete, and

suggested that both CH and C-S-H phases precipitate on their surface, consequently
leading to a more homogeneous distribution of hydration products. Most obvious
was the reduction in the amount and size of CH crystals at later ages. A study dealing
with the role of concrete incorporating high volume of fly ash in controlling the
expansion due to alkali aggregate reactions was reported by Malhotra and Alasali,
who indicated that the fly ash was highly effective in inhibiting the alkali-silica
reaction [45]. Similar results were presented by Malhotra et ai, and they indicated
very low chloride-ion diffusion in HVF A concrete at 91 days, ranging from 197 to
973 coulombs [63].
In 1990, CANMET undertook to develop an engineering data base on HVF A

concrete. Eight fly ashes, covering a wide range of mineralogical and chemical
compositions, and two portland cements from different sources were selected for the
study. The investigation was carried out for the Electric Power Research Institute
(EPRI), Palo Alto, California, under a subcontract with Radian Canada Inc. [9].
Briefly, in this HVF A concrete, the water and cement contents are kept low at about
115 and 155 kg/m3 of concrete, respectively, and the proportion of fly ash in the total
cementitious materials content ranged from 55 to 60 percent. The results have shown
excellent mechanical properties at both early and late ages, with compressive
strengths reaching as high as 50 MPa at 91 days. Values of Young's modulus of
elasticity of the order of 40 GPa at 91 days were achieved for the HVF A concretes
investigated. The creep and drying shrinkage of the HVF A concretes investigated
were relatively low.
Some research data on engineering properties from investigations of HVF A

concrete mixtures are presented in Table 2-5.
A mix of particular interest is Malhotra et al of the Table 2-5 with a water /

cementitious ratio of 0.30, and a total content of cementitious material of 340 kg/m 3 .
The compressive strength at 1 day was 0.6 MPa but it reached 79 MPa at the age of
1 year. It should be emphasized that experimental production of HVF A concrete
necessitates a very strict and consistent quality control.
35
36
Table 2-5 Mix proportions and engineering properties of some HVFA concretes
Ingredient Swamy & Mahmud Giaccio & Malhotra et al

(kg/m3 ) [59 ] Malhotra [601
[ 61 ]
Total cementitious 266 348 466 375.7 400 340 340 225 340
content
Portland cement 133 174 233 154 180 150 150 100 150
Fly ash 133 174 233 211 220 190 190 125 190
Silica fume -- - -- 10.7 -- -- - --
IWater / Cementitiom 0.615 0.420 0.320 0.31 0.28 0.30 0.33 0.49 0.35
SP % oftota1 1.7% 1.8% 1.8% 1.13% - - - -
cementitious
Slump ( mm) - - - - 120 200 100 100 210
!Compressive strengtl 100nun 100nun lOOnun 102x203 150x30C 150x30C 150x300 150x30C 150x300
(MPa) cubes cubes cubes nun mm mm rrun mm mm
at age ( days) cylinders cylinders o..,ylindeTI (;ylindcTI cylinders cylinders
1 4.4 11.8 21.8 9.3 13.7 0.6 10.2 4.0 2.7
7 13 31.3 43.3 17.6 34.2 26 25.8 12.1 20.5
28 23.7 46.6 63.3 32.5 57.1 49.1 46.1 23.0 37.5
91 46.8 70.0 88.8 41.2 75.2 63.0 61.8 37.9 53.0
365 49.5 75.6 92.5 -- -- 79.0 - - 69.0
Flexural strength 100 x 100 100 x 100 100 x 100 r6x102x40 150x30G 150x30C 150x30C 150x30C 150x300
(MPa) x500mm x500mm x500mm 6mm mm mm mm nun mm
at age ( days) pnsms prisms pnsms lCylinders o..,ylinders (;ylinders lCylindeTI cylinders
1 0.78 1.92 3.12 - -- - -- -- --
7 1.80 3.12 3.92 - - - -- - -
28 2.56 4.56 4.52 4.9 - 6.9 - - 6.0
91 4.88 6.64 7.64 - 9.6 - 8.9 - -

365 4.70 5.80 7.52 - - 9.0 - - 7.5
Modulus of e1asticit) 100xlOO 100 x 100 100 x 100 152x305m 150x30G 150x30C 150x30C 150x30C 150x300
(GPa) x500mm x500mm x500mm mn mm mm mm mm mm
at age ( days ) pnsms prisms pnsms cylinders lCylinders lCylindeTI lCylinders ~ylinders cylinders
28 35.6 41.7 46.9 35.5 36.1 35.1 33.8 27.9 -
Drying shrinkage 100 x 100 100 x 100 100 x 100 - - - - - -

( microstrain ) x500mm x500mm x500mm
at age ( days) pnsms prisms pnsms
7 169 163 187 - - - - - -
28 365 319 325 - - - - - -
91 430 405 413 - - - - - -
365 478 452 468 - - - - - --

730 485 460 475 - - - - - -
36
Table 2-5 ( Contd.! )
Ingredient V. Sivasundararn et al Langley et al Georges Carette et al

(kg/m3 ) [63 ] [27] r 431
Total cementitious 545 520 363 360 380 225 363 365 363
content
Portland cement 229 219 153 152 180 100 152 153 152
Fly ash 316 301 210 208 200 125 211 212 211
rvv ater / Cementitiom 0.21 0.22 0.31 0.32 0.27 0.49 0.33 0.33 0.33
Slump (mm) 200 200 200 200 200 110 140 110 190
~ompressi ve strengtl 152x305 152x305 152x305 152x305 152x305 152x305 152x305 152x305 152x305
(MPa) mm mm mm mm mm mm mm rnm mm
at age ( days ) cylinders cylinders cylinders cylinders cylinders ~inders cylinders cylinders cylinders
1 - - 2.8 - - - 5.6 10.3 5.8
7 - -- 17.8 15.6 27.5 - 18.8 24.9 18.2
28 55.4 48.2 28.2 27.2 44.0 22.1 28.8 39.3 27.8
91 -- - 35.9 32.9 59.7 35.5 39.7 50.9 39.5
365 - - 39.1 36.6 67.3 - 52.0 57.9 54.7
730 - - - - 73.3 - 58.4 65.2 55.0
Flexural strength -- - - -- 175x 100x [75x100x r5x 102x r5x102x [75 x 102x
(MPa) 400mm 400mm 406mm 406mm 406mm
at age ( days) prisms ~risms ~risms prisms prisms
14 --- -- -- - - - 3.7 4.8 3.4
91 - - - - 72 5.6 5.2 6.5 5.4
365 -- --- - - 7.5 6.3 -- --- ---
730 -- --- -- -- - -- 6.2 6.5 6.0
Modulus of elasticit) 152x305 152x305 152x305 152x305 152x305 152x305 152x305 152x305 152x305
(GPa) mm mm mm mm mm mm mm mm mm
at age ( days) cylinders cylinders cylinders cylinders cylinders cylinders cylinders cylinders cylinders
28 41 37.9 -- -- 30.1 27.9 33 35.8 32.8
91 -- - - - ;
(
- - 39.2 41.0 40.4
125 45.2 44.5 -- - -- -- - - -
365 - - - -- 41.3 - 42.4 42.6 43.6
Drying shrinkage -- - -- -- - - 76xl02x 76xl02x 176x102x
( microstrain) 390mm 390mm 390mm
at age ( days) prisms prisms prisms
7 - - - - - - 139 245 192
28 - - - - - - 341 422 326
56 - - - - - - 389 493 369
112 - - - - - 380 536 426
224 - - - - - 418 582 461
448 - - - -- - - 433 589 454
Chloride-ion penetration ~100x50 ~100x50 ~100x50 ~100x50 - - 102x203 102x203 102x203
( Charge in coulombs) mm mm mm mm mm mm mm
at age ( days) Icylinders cylinders cylinders
cores cores cores cores
7 - 8484 7821 - - - - -
28 --- 2130 935 - - 1804 815 1743
91 - - 539 - - 585 297 635
365 - - - - 136 127 156
Pulse Velocity - - - - 152x305 152x305 - - -
( kmlsec ) nUll mm
at age ( days) cylinders cylinders
7 - - - - 4.22 - - - -
42 - - - - 4.76 4.22 - - -
365 - - - - 4.78 4.47 - - -
730 - - - - 4.84 4.60 - - -
37
Chapter 3 Experimental Programme 38
CHAPTER THREE
EXPERIMENTAL PROGRAMME
3-1. INTRODUCTION
In order to understand the role of fly ash in concrete, and to develop a good method
of mixture proportioning of high volume fly ash concrete ( HVF A ), several mixtures
were tested and the results analysed. The development of strength in high volume fly
ash concrete is intimately related to mix proportioning, and the method of mix
proportioning will influence the properties of such concrete. The underlying principle
of the method used here is to produce high performance concrete for structural
applications. The first stage of this research consists in the development of several
mixtures of high volume fly ash concrete and the details are presented below.
,I
3-2. TEST PROGRAMME AND DETAILS OF CONCRETE

MIXTURES
Several methods of mixture proportioning of fly ash concrete have been developed
over the years. Basically, fly ash has been used as partial replacement of cement, fine
aggregate, or cement and fine aggregate. This method of mixture design has been
used extensively in many studies, and is also adopted for use in this research. The mix
parameters and proportions of the cement replacement mixtures are summarized in
Table 3-1 and Table 3-2.
All experimental work was performed in a laboratory environment, i.e., at

room temperature, and pressure, except when otherwise specified. A high range
38
water reducer ( fIR WR ) was used in all the mixes, and the dosage of this
superplasticizer was adjusted to give a slump of 150 '" 200 mm.
Table 3-1 High volume fly ash concrete test parameters
Parameter Range of variables

Type of mixture ( OPC Mixes) Mix-350C, Mix-450C
(HVFA Mixes) Mix-350F, Mix-350SF, Mix-350S

Mix-45 OF, Mix-450SF, Mix-450S
Total binder content 350 and 450
(kg / m3 )
Method of replacement Ie Partial replacement of cement
Ie Partial replacement of sand
Water / binder ratio 0.40
Size of specimen ( mm ) 100 x 100 x 500
Curing condition • Wet curing continuous in water
• 7 days water curing + continuous air curing
Slump (mm) ;
!I50 ~ 200
3
Table 3-2 High volume fly ash concrete mixtures (Unit: kg/m )
Type of mixture Total Cement Fly ash Silica fume

and cementitious
identification content
350C 350 350 --- ---
350F 350 150 200 ---
350SF 350 150 180 20
350S 350 150 180 20
450C 450 450 --- ---

450F 450 200 250 ---
450SF 450 200 230 20
450S 450 200 230 20

Remarks: Mix-350S and Mix-450S had both cement and sand replacement with fly
ash.
39
3-3. CHARACTERISTICS OF MATERIALS
Cement
The cement used was ordinary portland cement ( oPc ) conforming to British
Standard obtained from Castle Cement Limited, Lancashire. The specific surface area
of this cement was 345 m2/kg.
Fly ash
A Class F fly ash with the ASTM designation C618 was used in this
investigation. The fly ash used in this concrete was obtained from Boral Pozzolan
Ltd, Herts. The MgO content of l.62%, satisfies to the BS 3892 : Part 1 : 1981
"Pulverized-fuel ash for use as a cementitious component in structural concrete",
which restricts the MgO content to less than 4 %.
Silica fume
A silica fume slurry containing 50 % by weight of water was used. The silica
fume was supplied by Elkem Materials Limited in slurry form. Its specific gravity is
2.2. The chemical composition of the portland cement and mineral admixtures are
shown in Table 3-3.
High Range Water Reducer
A superplasticizer was used in this programme ; a concrete plasticiser P6

obtained from W. R. Grace Ltd. Cormix Division. This high range water reducer was
used to provide an effective means of producing low water-binder ratio concrete
combined with good workability. For the same workability of a concrete mix, the mix
incorporating a superplasticizer would need less water, leading to a lower
water/cementitious ratio, with a high workability.
Normal Aggregate
The fine and coarse aggregates were obtained from the Redlands quarry. A
sieve analysis was carried out on representative samples in accordance with BS 812 :
Part 103. 1: 1985 "Sieve tests". The sieve analysis results of the aggregates are given
.to
Chapter 3 Experimental Programme
in Table 3-4. The fine aggregate had a water absorption coefficient of 2.1 %, and a
bulk specific gravity 2.62, in saturated surface dry condition. The coarse aggregate
consisted of a mixture of rounded and crushed gravel with 10 nun maximum particle
size. The coarse aggregate water absorption was 0.3 % and bulk specific gravity
2.62, in saturated surface dry condition. Chemical analysis ( XRF method) of the
aggregates was provided by the aggregate supplier. The results of these tests are
shown in Table 3-5.
Table 3-3 Chemical composition of cements and mineral admixtures

(Unit: %)
Chemical composition OPC Fly ash Silica fume

Si02 20.4 51.4 97.0
AI203 5.70 28.1 -
Fe203 2.10 11.1 -
CaO 65.3 1.38 -
MgO 1.10 1.62 -
S03 3.20 - -
Na20 0.20 1.32 -
K20 0.59 4.08 -

Ti02 - 0.77 -
C3 S 54.3 - -
C2 S 17.6 - -
C3 A 11.6 - -
C4 AF 6.40 - -
Fe - - 0.09
AI - - 0.09
Ca - - 0.11
Mg - - 0.09
K - - 0.30
Na - - 0.09
H2 O - - 0.50
Loss of Ignition 1.00 3.50 1.10

2
Specific surface ( m / kg) 345 370 - 400 -
.tl
Chapter 3 !ixperimenfal Programme
Table 3-4 Sieve analysis of fine aggregate and coarse aggregate
Sieve size Fine aggregate Coarse aggregate

(nun) Passing (%) BS 882: (%) Passing (%) BS 882:( %)
13.2 100 100
9.5 96 85 - 100
4.75 20 0-25
2.36 3 0-5
9.5 100 100
4.75 93 89 - 100
2.36 84 60 - 100
1.18 80 30 - 100
0.6 77 15 - 100
0.3 39 5 -70
0.15 6 0-15
F.M. 2.21 1.81
Table 3-5 Chemical compositions of fine a~gregate and coarse aggregate
Chemical Content (% )
composition Fine aggregate Coarse aggregate
Si02 91.72 92.57
Ti02 0.13 0.25
Al203 3.86 2.22
Fe203 0.96 1.28
CaO 0.50 1.53
MgO 0.34 0.98
Na20 N/A 0.00
K20 2.18 0.93
Cr203 0.025 0.03
Mn304 0.005 0.009
P205 0.02 --
BaO 0.04 0.02
V205 0.007 --
Loss of Ignition 0.95 1.98
42
Chapter 3 Experimental Programme -13
3-4. CONCRETE MIXTURE DESIGN
The total cementitious content was kept constant in all the specimens at 350 kg/m3
3
and 450 kg/m respectively. The water / binder ratio of all the specimens was also
kept constant at 0.40. The aggregate / binder ratio was 5.2/ 1 for 350 kg/m3 mixes,
3
4.0411 for 450 kg/m mixes and the percentage of fine aggregate in the total was 35
% for the total aggregate. In the mixes studied in this project in addition to cement
replacement, a proportion of the fine aggregate, amounting to 25 %, was also
replaced by fly ash for Mix-350S and Mix-450S. The final mix proportions of high
volume fly ash concrete considered in this study are listed in Table 3-6.
Table 3-6 High volume fly ash concrete mix proportions ( Unit: kg/m3 )
Type of Total Cement Fly ash SF Water Coarse Fine

mixture binder aggregate aggregate
and content
identification
350C 350 350 --- --- 140 1183 637
350F 350 150 200 --- 140 1183 637
350SF 350 150 180 20 140 1183 637
350S 350 150 180 20 140 1183 160( Fly ash )
( 340)* 477( Sand)
450C 450 450 --- --- 180 1183 637
450F 450 200 250 --- 180 1183 637
450SF 450 200 230 20 180 1183 637
450S 450 200 230 20 180 1183 160( Fly ash)
(390 )* 477( Sand)
Remarks:
3
1. Total cementitious : Aggregate = 1 : 5.2 for 350 kg/m mixes, 1 : 4.04 for 450
kg/m3 mixes
2. 35 % Total aggregate = Fine aggregate
3. SF = Silica fume
4. Mix-350S and Mix-450S had both cement and sand replacement with fly ash
5. *: Total fly ash content
·0
Chapter 3 Experimental Programme 4-1
3-5. MANUFACTURE OF TEST SPECIMENS AND CURING
The fine and coarse aggregates were used in an air dry condition. Water content was
checked and adjusted before mixing. Due to very low water / binder ratio, different
methods of combining the constituents were tried and the following method was
found to obtain a homogenous mix and without bleeding.
1. The coarse aggregates were first placed in the mixer and half the quantity of
water was than added and the mixer was spun for about a minute.
2. The cement and mineral admixture with the fine aggregates were added, then
mixed 30 seconds.
3. Added remaining 1 / 2 of mixing water, then mixed for 90 seconds.
4. Superplasticizer was then added directly to the mixer and mixed for a further
minute.
5. Slumps of the mixes were performed after superplasticizer addition.
All specimens were cast in steel moulds which had been lightly oiled. After
casting, all the specimens were covered by polythene sheet and left in the laboratory.
The specimens were demoulded after 24 hours, and then cured in water for 7 days.
Curing was then continued either in water or in a drying environment. Eight batches
were cast for all the concrete mixtures. Each batch contained six prisms, three prisms
for water curing, and three for curing in a drying environment respectively.
Only one geometry of test specimen was used in the Part I of this research
study, namely, 100 x 100 x 500 mm prisms. The advantage of using 100 x 100 x 500
mm prisms test specimens for this part of the study is that with a single test geometry,
the essential properties of the concrete can be established, thus reducing the amount
of time and effort required for casting various sizes of specimens.
Chapter 3 Experimental Programme
45
3-6. DOSAGE OF SUPERPLASTICIZER ( HRWR ) AND SLUMP
From an engineering application point of view of the high volume fly ash concrete,
the mixture proportioning should produce a concrete with adequate flow
characteristics, so that the concrete can be easily placed and compacted. To achieve
this purpose, a relatively large variation in the superplasticizer dosage requirement
was needed, the quantities ranging from 0.9 to 2.5 percent of the total cementitious
content for the different concrete mixtures. The difference in the dosage of the
superplasticizer was mainly due to the different cementitious contents of the high
volume fly ash concrete mixes.
The workability in these studies was determined by the slump test. All the
mixes appeared to have good cohesiveness and no bleeding was detected. With
structural concrete, the aim is to produce a specified workability and 28 day normally
cured strength. The mixes in this study were designed to have a slump of 150 to 180
mm, and the results are shown in Table 3-7.
Table 3-7 High volume fly ash concrete workability
Type of mixture Total Superplasticizer Slump

and cementitious (%) of (mm)
identification content Cementitious
material
350C 350 2.2 165
350F 350 0.9 165
350SF 350 1.1 165
350S 350 1.6 165
450C 450 2.5 165
450F 450 1.2 160
450SF 450 1.3 165
450S 450 1.8 165
The normal portland cement concrete mixes had the highest demand of
HRWR of 2.2 % to 2.5 % of the cement content. Mix-350F had the most workable
concrete with the lowest dosage of the superplasticizer. The dosages required for
~5
concrete made with Mix-45 OF, Mix-350SF, and Mix-450SF were also quite low.
Mix-350S and Mix-450S had the highest water demand of all concretes with mineral
admixtures, and consequently, higher superplasticizer dosages of 1.6 % - 1.8 % of
total cementitious material were required. The high water demand of Mix-350S and
Mix-450S is most likely due to the higher fly ash content, and the presence of a high
proportion of fine material.
3-7. TEST DETAILS
The eight batches of specimens were used for determining the ultrasonic pulse
velocity, dynamic modulus of elasticity, drying shrinkage and expansion at various
ages up to 18 months. At each specific age, the batches were also used for the
determination of air permeability, water absorption, carbonation and compressive
strength following the flexural strength test. After the flexural strength test, each
broken prism was used for a modified cube compressive strength test. The
carbonation test followed the compressive strength test. For each test, three prisms
were used. Table 3-8 shows the details of the tests and the test specimen size used.
The tests are described in detail below.
3-7-1. Flexural and compressive strength test
One of the objectives of this study was to establish whether the incorporation of silica
fume in a high volume fly ash content was beneficial to the early strength
development. Accordingly, flexural and compressive strength tests were carried out
for all of the concretes at 1, 3, 7, 28, 90, 180, 360 days and 18 months after moist or
air curing. At each testing age, three specimens were tested ~ they were removed
from water one hour before testing. For flexural strength, the test procedure was
according to BS 1881 : Part 118 :1983 "Method of determination of flexural
strength" .
~6
Chapter 3 Experimental Programme -17
The compressive strength of the concrete was found by using portions of

prisms broken in the flexural tests ( equivalent cube method ). The equivalent cube
method of determining the compressive strength of concrete consists of applying the
load through 100 x 100 mm square auxiliary platens, equal to the nominal cross-
sectioned dimensions of the prism. The test was carried out according to BS 1881 :
Part 119 :1983, "Method for determination of compressive strength using portion of
beams broken in flexure".
3-7-2. Dynamic modulus of elasticity
The dynamic modulus of elasticity of concrete is related to the structural stiffness and
deformation process of concrete structures. The test was determined by the
"Resonant Frequency Tester", made by eNS Electronics Ltd. The instrument
consisted of an "Erudite" electromagnetic exciter unit, an electromagnetic pick-up
unit, a digital counter unit, a variable frequency oscillator and a cathode ray
oscilloscope. The test was conducted by placing the electromagnetic exciter unit and
the pick-up unit, along the longitudinal axis, at the centre of the specimen, against the
100 x 100 mm sides with a path length of 500 mm for each prism. Tests were
performed on the same specimens at intervals of 7 days. The test principle, procedure
and calculation of the value were all according to BS 1881 : Part 209 : 1990,
"Recommendations for the measurement of dynamic modulus of elasticity" .
3-7-3. Ultrasonic pulse velocity
The purpose of the ultrasonic pulse velocity tests was to look into the effects of the
two different curing conditions with time, and to monitor the development of the
internal structure and microcracking of the high volume fly ash concrete .
.t7
The ultrasonic pulse velocity tester used in the present study was the
PUNDIT, manufactured by C.N.S. Electronics Ltd. 54 kHz transducers of 50 mm
diameter were used. The measurements were taken at the two 100 x 100 mm sides,
along the longitudinal axis of the prisms. The measurement intervals at which the
ultrasonic pulse velocity was recorded were the same as for the dynamic modulus of
elasticity test. The test method was in accordance with BS 1881 : Part 203 : 1986,
"Recommendation for measurement of velocity of ultrasonic pulse in concrete".
Table 3-8 High volume fly ash concrete test details
Type of testing Specimen size Standard

(mm)
Flexural strength 100 x 100 x 500~ BS 1881 : Part 118 : 1983 "Method of
determination of flexural strength".
Modified cube 100 x 100 x 5001e BS 1881 : Part 119 : 1983 "Method for
compressive strength determination of compressive strength
using portion of beams broken in
flexure".
• BS 1881 : Part 116 : 1983 "Method for
determination of compressive strength of
concrete cubes".
Dynamic modulus of 100 x 100 x 500. BS 1881 :part209: 1990,
elasticity "Recommendations for the measurement
of dynamic modulus of elasticity" .
Ultrasonic pulse velocity 100 x 100 x 500 ~ BS 1881 : Part 203 : 1986,
"Recommendation for measurement of
velocity of ultrasonic pulse in concrete".
Swelling 100 x 100 x 500 I- BS 1881 : Part 5 : 1970, "Determination
of changes in length on drying and
wetting".
Drying shrinkage 100 x 100 x 500 • BS 6073 : Part 1 : 1981, "Specification
for precast concrete masonry units "
Appendix D " Determination of drying
shrinkage" .
Carbonation 100 x 100 x 500 Ie The depth of carbonation is measured
( Phenolphthalein Test ) using phenolphthalein solution.
"Measurement of hardened concrete
carbonation depth" [64].
Air permeability 50 x 30 cylinder I- "A new gas permeameter for measuring
the permeability of mortar and concrete".
Water absorption 75 xl 00 cylinder I- BS 1881 : Part 122 : 1983 "Method for
determination of water absorption" .
.t8
3-7-4. Swelling and drying shrinkage
The volume of a concrete varies with its water content. Drying will cause a volume
decrease, i.e., drying shrinkage, and wetting will cause a volume increase, i.e.,
swelling. Consequently, the rate of removal of water from concrete controls the rate
of shrinkage of the concrete. Reasonably, all factors which affect drying, for example
: temperature, humidity, pressure, air movement, etc., will actually control the
shrinkage.
To study the swelling and drying shrinkage characteristics of high volume fly
ash concrete, water curing and air curing were chosen. After demoulding, all the
specimens had two locating discs bonded to the concrete over a 200 mm gauge
length in the longitudinal direction in the middle of each of the four sides. Initial
readings were taken within an hour of fixing the discs. The measurements were taken
over a period of 18 months at regular intervals of 7 days. For each curing condition,
three specimens were tested, the results are thus the average of twelve readings. The
demec gauge can be read accurately to 8 x 10-6 strain. The swelling and drying
shrinkage were measured according to BS 6073 : Part 1 : 1981, "Specification for
precast concrete masonry units ", Appendix D, "Determination of drying shrinkage".
3-7-5. Carbonation
Carbon dioxide which penetrates the surface of concrete can react with alkaline
components in the concrete, mainly Ca(OH)2. This process leads to a reduction of
the pH value of the pore solution. The depth of carbonation is measured using
phenolphthalein solution [64].
At the specific age of the flexural strength test, the broken surface was
immediately cleaned of dust and loose particles after breaking. Carbonation tests
were carried out on the freshly broken surface by spraying with phenolphthalein
solution. The depth of the carbonation layer from the external surface of concrete to
49
the start of the coloration of the concrete was measured as depth of carbonation.
Eight measurements were made on each of the faces of each broken pris~ therefore,
the depth of carbonation reported is the average result of 24 readings.
50
Chapter 4 Test Results and Discussion of Part I 51
CHAPTER FOUR
TEST RESULTS AND DISCUSSION OF PART I
4-1. INTRODUCTION
A series of concrete mixtures involving two different curing conditions, with and
without fly ash and / or silica fume, with total cementitious contents of 3 50 kg/m3 and
3
450 kg/m , were designed. All the HVF A concrete mixes were made with a water /
binder ratio of 0.4. Because of the very low water content used, large dosages of a
superplasticizer had to be used to obtain the required workability.
In this chapter, the results of the effect of pozzolanic reaction on the

engineering properties of concrete such as compressive and flexural strength,
dynamics modulus of elasticity, ultrasonic pulse velocity, swelling and drying
shrinkage are reported at various ages up to 18 months. The effect of fly ash on
carbonation is also presented and discussed.
4-2. STRENGTH DEVELOPMENT
The strength of a concrete is its most important property because, after all, concrete
is a construction material. The hardening of the concrete is the result of the hydration
reaction, the subsequent development of bonds in the hydration products, and gradual
increase of the internal strength. Nevertheless, strength usually gives an overall
picture of the quality of concrete, because strength is directly related to the structure
of the hydrated cement paste. Moreover, the strength of concrete is almost invariably
a vital element of structural design, and is specified for compliance purposes.
51
Chapter 4 Test Results and Discussion o[Part J 52
This section presents the results of the strength properties of high volume fly
ash concrete in its hardened form. For purpose of comparison, the strength properties
of normal portland cement concrete are also discussed .
The flexural and compressive strength tests were carried out for all of the
concretes at 1, 3, 7, 28, 90, 360 days and 18 months of moist curing and air curing.
Three specimens were tested at each age, and the results are thus the average of three
measurements. The compressive strength of the concrete was found by the modified
cube test.
4-2-1. Effect of curing condition on compressive strength
The data on the compressive strength development under both curing conditions are
presented in Tables 4-1 and 4-3 and illustrated in Figs 4-1 and 4-3.
The compressive strength expressed as a percentage of the 28 day values

under both curing conditions is given in Tables 4-2 and 4-4, and Figs 4-2 and 4-4.
The strength development pattern is quite different for the two curing conditions, and
cementitious contents, and these are therefore discussed separately.
Water curing on compressive strength development
Early strength development is a matter of great importance to the

construction industry, and closely related both to the performance and economics of
concrete at early ages [6].
The compressive strength development of high volume fly ash concrete under
water curing is shown in Table 4-1 and compared in Fig 4-1.
It can be seen from the table that, the early age compressive strength of
concrete incorporating both fly ash and silica fume, i.e. of, Mix-350SF and Mix-
450SF, is higher in comparison with those of concrete with cement replaced partially
by fly ash alone i.e., Mix-350F, Mix-450F, but lower than the strength of mixes
52
containing fly ash as partial replacement of both sand and cement. The increased gain
in strength at early ages of concrete incorporating fly ash and silica fume may suggest
an early start of the pozzolanic reaction due to the use of silica fume, although this
may also be partially accounted for by the filler effect of silica fume resulting in a
denser matrix structure [65].
The data in Table 4-1 show that the compressive strength generally increased
with age for all the HVFA concretes, producing high compressive strength in the
range of 60 to 68 MPa at the age of 18 months. The data also show that the presence
of silica fume had very little effect on the long-term strength of the HVF A concrete.
There are two possible reasons for the higher strength of Mix-350SF and
Mix-450SF concretes at early ages. First, because of its small particle size, silica fume
can act as a filler in the spaces between cement grains. This results in a reduction in
the size of the individual pores and voids in the concrete, although the total porosity
is not affected. Since pores are discontinuities in the cement concrete matrix, reduced
pore sizes require a higher stress to initiate a crack ; thus, the strength is increased
[54]. A second reason for the early high strength of the 350SF, 450SF concrete is
due to the pozzolanic nature of silica fume, which also explains why the 350SF,
450SF concretes become progressively stronger with time relative to the concretes
containing fly ash only.
Mix-350S and Mix-450S containing fly ash as both cement and sand
replacement showed slightly higher strength at all ages than the mixes without and
with silica fume. As mentioned before, early fly ash concrete mixes were based on
low percentage of cement replacements ; these methods are suitable for mass
concrete where early strength development is not a criterion, and they invariably lead
to a lower strength of fly ash concrete [6]. When fly ash is also used to replace part
of the sand in a mix without changing the water and cement contents, the total
cementitious material actually increases and the water / binder ( w / b ) ratio
decreases accordingly. Theoretically, the strength of concrete increases with
decreasing w / b ratio. Therefore, proper use of fly ash to partially replace sand will
be advantageous in strength development. Since the fly ash particle is smaller than the
53
sand, the more the sand is replaced, the more the specific surface area is increased.
However, if an excessive portion of sand is replaced by fly ash, the w / b ratio is also
excessively reduced, and thus may, in turn, create workability problems.
When fly ash is used to replace part of the cement and sand simultaneously,
the advantage gained from reducing the w / b ratio may be able to compensate the
loss in early strength due to the presence of fly ash. In other words, using fly ash to
replace both cement and sand partially appears to have a synergistic effect combining
the benefits of both separate replacements.
The results in Table 4-1 also show that the difference in compressive strength
3
between 350 kg/m mixes and 450 kg/m3 mixes is small. Both mixes give 12 to 20
MPa at 3 days, 20 to 30 MPa at 7 days, and 30 to 40 MPa at 28 days. At one year,
the compressive strength ranged from 55 to 60 MPa, and in excess of 60 MPa at 18
months.
Table 4-2 and Fig 4-2 show the rate of gain of compressive strength as a
percentage of the 28 day strength. The results confirm the role of pozzolanic
reaction. At 7 days, compressive strength values ranged from 60 % to 75 % for the
350 kg/m3 concretes, and from 76 % to 81 % for the 450 kg/m3 concretes of their 28
day strength. This was followed by a larger increment after long term water curing.
This increment at 18 months ranged from 157 % to 187 % for the 350 kg/m3 mixes
and from 163 % to 201 % for the 450 kg/m3 concretes compared to the 28 day
strength. These data show that, as expected, the strength development, in general,
increases with age for all the concrete mixtures. The data show that for moist cured
specimens, the higher the compressive strength, the greater was the increase in
strength up to 28 days. The percentage gain in strength of the higher strength
concrete was greater than that of the low strength concrete. However, this is
reversed after about six months ; the stronger the concrete the lower is its subsequent
proportionate gain in moist cured concrete. These results show that when curing
conditions are favourable, the pozzolanic reaction of fly ash continues over a long
period, with the contribution of pozzolanic reaction being especially more prominent
for flexural and tensile strength than for compressive strength [19].
54
Chapter 4 Test Results and Discussion ofPart 1 55
Table 4-1 Compressive strength ofBVFA concrete underwater curing

(Unit: MPa)
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 6.2 6.5 7.2 6.4 6.7 7.1
3 12.3 15.0 19.3 14.5 16.3 19.1
7 20.0 27.8 31.1 24.1 29.7 32.1
28 33.2 37.1 41.4 31.0 39.1 39.5
90 42.7 47.9 61.6 43.9 55.5 60.0
180 49.3 61.4 59.4 53.5 61.7 65.2
360 55.0 58.4 61.9 57.5 59.3 63.0
540 62.2 60.1 64.8 62.4 63.8 67.5
80
o:!
350F 350SF 350S
~ ~ --v- ---B- ~-'7~
,g 60 y-
~ -------~'
~
r.I.l 40
~
'Vi
'"
<I.)
6.20
~
U
'0 180 360 540
1 7 28 90
Age (Days)
80
o:!
450F 450SF 450S
~ -+-- ------ ---*-
·60
~
5
.......
r.I.l
<I.)
40
;>
"Vi
~
6.20
E
0 ~-.;~~;;:-~~~---
U
'0 28 90 180 360 540
1 7
Age ( Days)
Figure 4-1 Compressive strength ofBVFA concrete-Water curing
55
Chapter 4 Test Results and Discussion of Part 1 56

( Water curing)
(Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

(days)
1 19 18 17 20 17 18
3 37 40 47 47 42 48
7 60 75 75 78 76 81
28 100 100 100 100 100 100
90 129 129 149 142 142 152
180 148 165 143 173 158 165
360 166 157 150 185 152 159
540 187 162 157 201 163 171
200r-------------~========~------------~
350C 350F 350SF 350S
160 .....-- ~ ~ -e--
.,._...-4----
120 ._------_...__._-_........ ...
~
7 28 90 180 360
Age (Days)
200~----------~==========~------------~
450C 450F 450SF 450S
160 - - -+- ----- ------
120
80
40
O'L---~----~~~~------~------~~--~--~~
1 7 28 90 180 360
Age (Days)
Figure 4-2 Rate of compressive strength as % of 28 days strength-Water curing
56
Fig 4-2 also shows that the 28 day strength appears to be some sort of a
threshold stage. Before that mixes with pozzolanic materials appear to give lower
strength than the control concrete, whereas beyond 28 days, the control concrete
gives lower strength than the concrete with the pozzolanic materials.
Air curing on compressive strength development
Theoretically, curing is not generally considered as part of concrete mix

design, but with mineral admixtures that depend on their reactivity to contribute to
strength and durability, curing should be regarded as an essential ingredient of the
mix design [6]. The strength development with age for concretes moist cured 7 days
followed by air curing are shown in Table 4-3 and Fig 4-3.
Under air curing, the compressive strength development of all the concretes
show a similar pattern to that previously observed with water curing, the strength
increasing with increasing age for all the concrete mixtures. However, strength
development is generally lower than that under water curing. All concrete mixtures
had 28 to 40 MPa at 28 days compared to 33 to 41 MPa under water curing. Further,
the data show that for all air cured specimens there is a gradual retrogression of
compressive strength with continued air drying. At one year, the compressive
strength of all air cured concretes was 47 to 62 MPa compared to 55 to 63 MPa for
water curing. All the mixes except Mix-350S and Mix-450S generally showed lower
strength than that water cured specimens. Surprisingly, Mix-350S, Mix-450S showed
least loss of strength due to air drying. Both mixes gave about 57 ~ 58 MPa at 18
months compared to 65 ~ 68 MPa under water curing.
The rate of compressive strength development as a percentage of 28 days

strength under air curing is shown in Table 4-4 and illustrated in Fig 4-4. The results
show that the 7 day compressive strength values ranged from 70 % to 90 % for the
3
350 kg/m3 concretes and from 78 % to 84 % for the 450 kg/m concretes of their 28
days strength. These percentage values are higher than that the rate of compressive
strength development under water curing.
57
Table 4-3 Compressive strength of HVFA concrete under air curing

(Unit: MPa)
Age 350F 3508F 3508 450F 4508F 4508

( days)
1 6.2 6.5 7.2 6.4 6.7 7.1
3 12.3 15.0 19.3 14.5 16.3 19.1
7 20.0 27.8 31.1 24.1 29.7 32.1
28 28.4 29.9 38.9 28.6 38.1 39.3
90 41.0 46.0 56.3 42.7 56.0 58.5
180 42.0 56.5 61.9 51.7 54.6 63.0
360 46.8 54.8 60.9 47.0 55.7 62.0
540 47.1 52.6 58.2 45.9 56.4 57.2
80
~ 350F 350SF 350S
::E ~~-B-
,5 60
/)J)
~
r.I.l 40
(\)
;;.
'm
'"
(\)
ts. 20
E
0
C)
'0
1 7 28 90 180 360 540
Age (Days)
80
d
450F 450SF 450S
~ --t-- - - -
,..::' 60
bo
....~
r.I.l 40
(\)
,~
'"'"
(\)
ts. 20
0
E
C)
'0 90 180
1 7 28 360 540
Age (Days)
Figure 4-3 Compressive strength of HVFA concrete-Air curing
58

(Air curing)
( Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 22 22 18 22 18 18
3 43 50 50 51 43 49
7 70 93 80 84 78 82
28 100 100 100 100 100 100
90 144 154 145 149 147 149
180 148 189 159 181 143 160
360 165 183 157 164 146 158
540 166 176 150 160 148 146
200
350C 350F 350SF 350S
160 ._....... --*- -v- -e-
120
?fl
80
40
0 180 360
1 7 28 90
Age (Days)
200
450C 450F 450SF 450S
160 --+- - -
120
~
80
40
0' 28 90 180
1 7 360
Age (Days)
Figure 4-4 Rate of compressive strength as % of 28 days strength-Air curing
59
At one year and 18 months the rate of gain of strength under air curing was
less than that under water curing
The relative rates of development of strength of HVF A concretes cured in

water and in air are illustrated in Tables 4-5 and 4-6.
Table 4-5 Influence of air curing on compressive strength-350 kglm3 binder

content
(Unit: % )
Mix 350F 350SF 350S

Age Water Air Water Air Water Air
( days) curing curing curing curing curin~ curing
28 100 85.5 100 80.6 100 94.0
90 100 96.0 100 96.0 100 91.4
180 100 85.2 100 91.7 100 104.2
360 100 85.1 100 93.8 100 98.4
540 100 75.7 100 87.5 100 89.8
3
content
(Unit: % )
Mix 450F 450SF 450S

( days) curing curing curing curin~ curing curing
28 100 92.3 100 97.4 100 99.5
90 100 97.3 100 100.9 100 97.5
180 100 96.6 100 88.5 100 96.6
360 100 81.6 100 93.9 100 98.4
540 100 73.6 100 88.4 100 84.7
60
These results emphasize the effect of air curing in the long term. Air curing of
all concrete mixtures exhibits a slight retrogression of strength with continued air
drying. The average compressive strength of air cured concrete at 18 months varied
3
from 76 % to 90 % for the 350 kglm mixes, and from 74 % to 88 % for the 450
kg/m3 mixes of the water cured strength. It is recognized that high volume fly ash
concretes are sensitive to moisture conditions.
The results indicate that the best development of strength gain occurs in moist
curing rather than in air curing. The results also indicate that for a steady increase in
strength for concrete incorporating high volume fly ash content a continuous moist
environment is indispensable. In concretes with low water / cement ratios, once
enough moisture is lost from the concrete, so that the internal relative humidity drops
below about 80 % either by evaporation or self-desiccation, hydration will stop and
strength development will be reduced below its potential, and this reduction will be
greater in the case of high-strength concrete ( low water / cement ratio ) than it is in
the case of low-strength concrete ( high water / cement ratio ) [13]. The loss of
potential strength in the long term is a more widespread and insidious problem, but it
is avoidable. Therefore, it is desirable to provide additional moisture during curing to
ensure maximum hydration.
Concrete can be damaged by lack of proper curing. It is essential for the

development of high strengths that the concrete should be kept moist for as a long
period possible, and not allowed to dry out , until the concrete strength has attained
its specific strength. Obviously, this is seldom a practical proposition, and some loss
in strength is inevitable to allow acceptable construction schedules and to minimise
cost.
The ACI Recommended Practice for moist cunng for most structural
concrete, is the time necessary to attain 70 % of the specified compressive or flexural
strength, whichever is less. Various investigators have concluded that the percentage
of strength of dry cured concrete to that of water cured concrete is between 60 % to
90 %, or sometime this varies even more widely.
61
Chapter 4 Test Results and Discussion o[Part I 62
The results from the present investigation of all concrete mixtures under air
curing gave strength values from 70 % to 90 % to that under water curing at 18
months. These data emphasize dramatically the need for early incorporation and a
minimum period of water curing as an integral part of mix proportioning fly ash
concrete.
4-2-2. Effect of curing condition on flexural strength
The data on flexural strength development are presented in Tables 4-7, 4-9, and
illustrated in Figs 4-5 and 4-7. The flexural strength expressed as a percentage of 28
day values are given in Table 4-8 and 4-10 and in Figs 4-6 and 4-8.
These results show that the development of flexural strength occurs in a

similar way to that of the compressive strength for moist and dry cured specimens.
As for compressive strength, air cured concrete showed lower flexural strengths
compared to concrete subjected to continuous moist curing. At 18 months, flexural
strength under moist curing varied from 5.7 to 6.4 MPa for the 350 kglm3 mixes
compared to 5.5 to 6.5 MPa for air cured specimens. Similarly, the flexural strength
of 450 kglm3 mixes under water curing varied from 5.8 to 6.6 MPa compared to 5.8
MPa for air cured concrete.
The rates of strength development in Tables 4-8 and 4-10 show similar
patterns to compressive strength. Although the higher strength concrete had lower
percentage gain, the flexural strength value in the stronger concrete was higher than
that in the low strength concrete.
The relative rates of development of strength of high volume fly ash concrete
cured in water and in air are illustrated in Tables 4-11 and 4-12. At the end of 18
months, all concrete mixtures exhibited a very minimal reduction in flexural strength
with continued air drying, i.e. about 97 % of the average strength for water cured
concrete. Except Mix-450S which showed a higher loss.
62
Table 4-7 Flexural strength of HVFA concrete under water curing

(Unit: MPa)
Age 350F 3508F 3508 450F 450SF 4508

( days)
1 1.30 1.79 1.85 1.33 1.81 1.87
3 2.25 3.04 3.27 2.23 3.07 3.25
7 2.78 3.07 3.70 2.87 3.15 3.87
28 3.48 4.52 4.78 3.56 4.61 4.89
90 5.07 5.51 5.83 5.19 5.65 6.54
180 5.25 5.69 6.65 5.46 6.51 6.23
360 5.55 5.71 5.85 5.67 5.75 6.63
540 5.70 5.74 6.36 5.76 5.88 6.60
~0'6
;;.
'/il
o_~
e'"
S'4
8
12 ----~-- 350F 350SF 350S
CI)
G: -*- ~ -&-
'0
1 7 28 90 180 360 540
Age (Days)
8
<IS
p..
~
0'6
.::
'"'"
CI)
....
e0..4
0
u
"i
.... 2
450F 450SF 450S
~ -+-- - ----*-
G:
'0
1 7 28 90 180 360 540
Age (Days)
Figure 4-5 Flexural strength of HVFA concrete-Water curing
63
Chapter 4 Test Results and Discussion o(Part 1 64
Table 4-8 Rate of gain of flexural strength as a 010 of 28 day strength

( Water curing)
( Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

(days)
1 37 40 39 37 39 38
3 65 67 68 63 67 66
7 80 68 77 81 68 79
28 100 100 100 100 100 100
90 146 122 122 146 123 134
180 151 126 139 153 141 127
360 159 126 122 159 125 136
540 163 127 133 162 128 135
200
350C 350F 350SF 350S
160 .-.-..• ~ ~ ----e-
120
--.--.
...~...........-.-.-..
~
80 _....-----------------_.
40
0 90 180 360
1 7 28
Age (Days)
200
450C 450F 450SF 450S
160 --r- - ---*-
120
~
80
40
0' 28 90 180 360
1 7
Age (Days)
Figure 4-6 Rate of flexural strength as % of 28 days strength-Water curing
64
Table 4-9 Flexural strength of HVFA concrete under air curing

(Unit: MPa)
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 1.30 1.79 1.85 1.33 1.81 1.87
3 2.25 3.04 3.27 2.23 3.07 3.25
7 2.78 3.07 3.70 2.87 3.15 3.87
28 3.17 3.18 3.89 3.23 3.29 3.95
90 5.03 5.15 6.18 5.13 5.22 6.36
180 4.48 5.22 5.91 4.58 5.67 6.09
360 4.89 4.89 5.70 4.91 5.19 6.14
540 5.49 5.52 6.54 5.85 5.76 5.79
8~-------------------------------------------------'
350F 350SF 350S

--fr-- ~ -----e-
7 28 90 180 360 540

Age ( Days)
8~-------------------------------------------------r
0:1
p..
:E
Q'6
;;.
"1jl
e'"
~4
U
til
.... 2 450F 450SF 450S
~ -+- -------- ----+-
G:
'0 L_~~~~--.l-...~-===:==;::::::===~~:-----=-:-:-~:-:-::-----:-:-:~
1 7 28 90 180 360 540
Age (Days)
Figure 4-7 Flexural strength of HVFA concrete-Air curing
65
Chapter 4 Test Results and Discussion o(Part I 66
Table 4-10 Rate of gain of flexural strength as a % of28 days strength

( Air curing)
( Unit: 0/0)
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 32 56 48 41 55 47
3 71 96 84 69 93 82
7 88 97 95 89 96 98
28 100 100 100 100 100 100
90 159 162 159 159 159 161
180 141 164 152 142 172 154
360 154 154 147 152 158 155
540 173 174 168 181 180 147
200
350C 350F 350SF 350S
160 _..._.- ~ -4- -B-
120 -...--.-
...
_-_ __
..... ... .--.. .... _--_.----
~
80
40
0 28 90 180 360
1 7
Age (Days)
200
450C 450F 450SF 450S
160
120
~
80
~--
40
0' 28 90 180 360
1 7
Age (Days)
Figure 4-8 Rate of flexural strength as % of 28 days strength-Air curing
66
Chapter 4 Test Results and Discussion o{Part 1 67
Table 4-11 Influence of air curing on flexural strength-350 kglm3 binder content
(Unit: %)
Mix 350F 350SF 350S

( days) curing curing curing curing curing curing
28 100 91.1 100 70.4 100 81.4
90 100 99.2 100 93.5 100 106
180 100 85.3 100 91.7 100 88.9
360 100 88.1 100 85.6 100 97.4
540 100 96.3 100 96.2 100 102
(Unit: % )
Mix 450F 450SF 450S

( days) curing curing curing curing curing curing
28 100 90.7 100 71.4 100 81.8
90 100 98.8 100 92.4 100 97.2
180 100 83.9 100 87.1 100 97.8
360 100 86.6 100 90.3 100 92.6
540 100 101.6 100 98.0 100 87.7
These results thus show that the decrease in flexural strength of air cured
concretes is relatively small, and almost negligible. The results indicate that air curing
preceded by moist curing, even for a short period, seems to have beneficial effects on
strength development. These results are in accordance with those reported by Swamy
who studied the effect of initial 7 day water curing on fly ash concrete, and indicated
that early water curing for a minimum period of seven days is an essential ingredient
67
for the successful use and full realisation of the strength potential of fly ash concrete
[6].
4-2-3. Relationship between flexural strength and compressive

strength
The relationship between flexural and the compressive strength is generally of the
form: [13]
f =a fc b -------------------------- ( 4-1 )
where f is the flexural strength in MPa, fc is the compressive strength in MPa, and a
and b are constants.
The variation of flexural strength with compreSSIve strength, for HVF A

concrete is shown in Figs 4-9 and 4-10 for water and air curing respectively. The
relationship shown is based on results tested at 1 day to 18 months for moist cured
and dry cured HVFA concrete.
Regression analysis defining the relationship between the two strength

properties gave as follows :
f= 0.52 fc 0.59 (moist) (r = 0.98 ) ---------- ( 4-2 )
f= 0.53 fc 0.58 (dry) ( r = 0.96 ) ---------- ( 4-3 )
where r is the correlation factor.
The correlation factor for the two sets of equations is high, and hence, in
practice, a single equation describing the two strength parameters can be used as
given below and shown in Fig 4-11.
f= 0.52 fc 0.59 (moist + dry) (r = 0.97 ) ---------- ( 4-4 )
68
Chapter 4 Test Results and Discussion o[Part I
10
350F 350SF 350S 450F 450SF 450S
X <> 0 + • I
~
8
~
~
~ .. ~
ts:: 6
. '*
0
x
+~~.
~
VJ
MOIST 0
-a \
3~ 4 •
~x
......
(])
o· o·
~ +
X
+
2
o ~~ __ ~~ __ ~~ __ ~~ __ ~~ __ ~~ __ ~~ __ __
~ L-~
o 10 20 30 40 50 60 70 80
Compressive Strength, MPa

HVFA concrete-Water curing
10 ~------------------------------------------~
350F 350SF 350S 450F 450SF 450S

X <> o + •
8
o
o • o
••
o
x+
DRY
\ o
2
+
/*
o L-~~ __~-L~__~~~--~-L--~~~--~~~
o 10 20 30 40 50 60 70 80

HVFA concrete-Air curing
69
These data indicate a significant correlation between flexural strength and

compressive strength of HVF A concretes at the 99 % confidence leveL In other
words, as the compressive strength, fe, increases, the flexural strength, f, also
increases.
Fig 4-11 is also compared to the data suggested by Swamy & Mahmud and
the relationship obtained from several other authors [9,27,59,60]. Regression analysis
of the data shown in the Fig 4-11 produced the following relationship :
f= 0.31 fe 0.71 (moist + dry) (r = 0.97) ---- ( 4-5) Swamy & Mahmud [59]
f= 1. 71 fc 0.33 (moist) (r = 0.64 ) ---- ( 4-6) EPRI [9]
f= 1.18 fe 0.46 (moist) (r = 0.80) ---- (4-7) Langleyet al [60]
f= 1.17 fe 0.44 (moist) ( r = 0.95 ) ---- ( 4-8) Malhotra et al [27]
10 ~------------------------------------------~
~~~- Swamy
Swamy EPR! Langley et al Malhotra et al Author

e -------0------- -----8----- -- ..... - -
0
0 10 20 30 40 50 60 70 80 90 100

HVFA concrete
70
Chapter 4 Test Results and Discussion orpar! I 7J
The results in Fig 4-11 show a wide variation in strength development ~ this is
mainly due to the way the HVF A concretes are proportioned, differences in initial
and type of curing, water / binder ratio, and age of testing. It can be observed from
the figure that the author's results do not diverge greatly from a single relationship.
Also, it is very comforting to see, a close similarity to the HVF A concrete data
obtained from Swamy & Mahmud.
Regression analysis showed that the coefficient, "r", usmg strength

development data in this relationship, was high for Swamy's and author's equations;
they are higher than those given by the EPRI and Langley et al. Due to different test
variables, it is difficult to interpret these results, and give definitive conclusions.
However, it is obvious from the data that concrete having a high cementitious
content, low water content and adequate curing, are most likely to possess a well-
developed strength.
4-2-4. Comparison with normal OPC concrete
The strength characteristics of HVFA concretes can be clearly inferred from Tables
4-1,4-3,4-7, and 4-9. The results also indicate the need for a well-defmed period of
water curing for the strength development of such concrete. It would then be
interesting to compare the strength development of HVF A concretes with that of
normal portland concrete at different ages, and under different curing conditions.
Table 4-13 represents the strength data for both ope concretes under both
types of curing conditions. The rate of ope concrete strength development as a
percentage of 28 day strength development compared to all HVF A concretes is also
shown in Figs 4-2, 4-4, 4-6, and 4-8.
Irrespective of the curing conditions, the rate of strength development in both

compression and flexure for both ope concretes is higher than that for HVF A
71
concretes at an early age, but after 28 days, it is lower than that for HVF A concretes
under the same conditions.
Table 4-13 Strength development of normal ope concrete

A. 350 kg/m3 mixes
Strength Compressive 0/0 of28 day Flexural % of28 day
strength strength strength strength
Age Water Air Water Air Water Air Water Air
( days) curing curing curing curing curing curing curing curin~
1 18.0 18.0 29 25 2.56 2.56 39 38
7 60.0 60.0 97 84 6.59 6.59 99 99
28 61.9 71.0 100 100 6.63 6.66 100 100
90 62.9 83.0 102 117 8.80 8.09 133 121
180 79.4 78.8 128 111 6.48 7.77 98 117
360 80.0 80.6 129 114 5.25 7.98 79 120
B. 450 kg/m3 mixes

Strength Compressive % of28 day Flexural % of28 day
strength strength strength strength
( days) curing curing curing curing curing curin~ curing curin~
1 20.2 20.2 31 29 2.69 2.69 42 50
7 59.5 59.5 90 84 6.32 6.32 98 117
28 66.2 70.5 100 100 6.42 5.40 100 100
90 70.0 83.2 106 118 7.19 7.80 112 144
180 75.2 76.5 114 109 6.83 7.88 106 146
360 77.7 77.1 117 109 6.03 8.19 94 152
Remark : Compressive and flexural strength values are given in MPa
From the results, it can be seen that under continued moist curing condition,
the percentage increase in compressive strength from 28 days to one year is 17 % ~
72
Chapter 4 Test Results and Discussion ofPart I 73
29 % for all normal OPC concretes. The corresponding increase in compressive

strength for the HVFA concretes was 50 % ~ 85 % ( Table 4-2 ). At the end of one
year, the percentage gain of flexural strength of OPC concrete was only 79 % ~ 94 %
of the 28 day flexural strength. The corresponding percentage gain in flexural
strength for the HVF A concretes was 122 % ~ 159 % of the 28 day flexural strength
( Table 4-8 ). The normal OPC concrete exhibited a slight degradation in flexural
strength even under continuous moist curing. For the concrete Mix-45 OF, the
percentage of compressive strength increase was more rapid after 90 days. Under
water curing, the difference between 450F and 450C on compressive or flexural
strength at one year was about 65 % respectively. These observations indicate the
potential for strength growth at later ages for fly ash concretes in constant contact
with moisture.
The results also indicate that under air curing, some loss of strength occurs
with OPC concretes with age, whereas HVF A concretes are more likely to maintain
hydration process with a gradual rate of increase. At the end of one year, a slight
retrogression of strength on HVF A concretes with air curing was similar to that of
normal OPC concrete, but the percentage gain on HVFA concrete was still higher.
For example, the difference between 350SF and 350C, 450SF and 450C at one year
with air curing was 69 % and 37 % on compressive strength, and 34 % and 6 % on
flexural strength, respectively.
These results indicate that even under continuous air drying, the HVF A
concretes are able to continue hydration process, although at a slower rate.
The superiority in the rate of later age strength development of the HVFA
concretes is clearly seen, even though had lower strength than that of OPC concretes.
This increase in the rate of strength development of the HVF A concretes over that of
OPC concretes was about 40 % of the average values on compressive strength, and
30 % on the flexural strength. The incorporation of fly ash, silica fume and a
superplasticizer in HVF A concrete is thus clearly beneficial to its long term strength
development.
73
Chapter 4 Test Results and Discussion of Part 1 7-1
4-2-5. Comparison with total binder content
The effect of total binder content on concrete strength development under both types
of curing conditions is shown in Figs 4-12, 4-13, 4-14, and 4-15. The results suggest
several aspects related to the role played by the superplasticizer and the binder
content. For purposes of comparison, the results for the normal portland cement
concrete are also included in the above figures.
The results show that the total cementitious binder content does not have
much effect on the value of strength. In general, concrete with 450 kg/m3 is only
3
marginally better than 350 kg/m mixes under both types of curing conditions.
In the current study, the compressive strength of 450C under both curing
conditions was somewhat lower than that of the concrete made of 350C. Previous
discussion has shown that the concretes of 450C had superplasticizer demand of2.5
%, higher than that for the concrete of 350C. The lower strength of concrete with
increasing dosage of superplasticizer is due, at least in part, to the greater fluidity and
the accompanying segregation and bleeding that occur in these concretes. These
results conform to the explanation of Goldman and Bentur [66] who used
progressively more superplasticizer as the silica fume content of their concrete was
increased. It appears that, although the workability of their concrete was enhanced,
the concretes became too fluid at the higher superplasticizer contents, resulting in a
reduction in strength similar to that observed in this study.
The results stress two points : that for the HVF A concrete to have a certain
specified strength, it is important to choose a suitable cementitious binder content
with the object of producing, as economically as possible, concrete of the required
strength. In addition, the selection of mix proportions has to take into account the
method of transporting the HVFA concrete, especially if pumping is envisaged.
74
Chapter 4 Test Results and Discussion o{Part I
100 100
'" '"
~ 80 ~ 80
350F450F
+
-!c oS
bJ)
:::
*
60 E 60
E
r/.l r/.l
u u
;> ;>
"!iJ 40 "!iJ
e'" '"
8
a- 20
0
1 7
I 35OC450C
28
I
90 540
! 7 28 90 540
Age ( Days) Age (Days)
100 100
~'"
<IS
350SF450SF 350S450S
~ 80 ~ 80 -e- .....
~g ~ :::
60 u 60
r/.l -=
r/.l
u u
;> ;>
"1jJ 40 "Til 40
e'" e'"0.
a-
0
80 20
U U
0
7 28 90 540 1 7 28 90 540
Age (Days) Age (Days)

( Water Curing)
100 100
o:s <IS
350F 450F
~ 80 ~ 80 +
tc ~c *
g 60 g 60
r/.l r/.l
u <I)
;> ;>
"1jJ 40 "a=J 40
en
e'" e0.
a- 20 I 350C450C I 8
u0 u0
0 7 28 90 540
1 7 28 90 540
100 100
o:s o:s
350S 450S
~ 350SF450SF ~
t:::
80 ~ --- ~
80 -e- .....
:::
u 60 <I)
60
-~
-=
r/.l -=
r/.l
<I) <I)
;>
;>
"1jJ 40 "1jJ 40
en en
e /~/ e0.
S' 20 ~/
80 20
U
0
~..v' u
0
7 28 90 540 1 7 28 90 540

( Air curing)
75
Chapter 4 Test Results and Discussion orPart 1
10 10
/\ = 350F450F
~ 8
/ ../
/"" \ . ~
..g
8
* +
..g
gp 6 /;l()
6
g g=
00 00
4 4
e 1~
~
~
(I)
2 I 350C450C I ~ 2
'0 0
1 7 28 90 540 1 7 28 90 540
10 10
3508F4508F = 35084508
~ 8 ......
~
~
~ 8 -e-
~g 6 ~g 6
00 00
e 4
e::1
4
~ 2
~
~ 2
~
'0 0
1 7 28 90 540 1 7 28 90 540
Figure 4-14 Effect of total binder content on flexural strength (Water curing)
10 10
350F 450F
~ 8 =
0... 8 +
~ ~ *
~ 6 ~ ~
6
.....~ .....
00
00
4
e 4
1~
~ 2 I 350C45OC I ~ 2
~
'0 0
1 7 28 90 540 1 7 28 90 540
10 10
35084508
350SF4508F
~ ....... =8
0... -e- ......
~
~~
~ 6
J''<:' • • .....~
00
4
:<V
-"'" ,-~
1~ 2
'0 0
1 7 28 90 540 1 7 28 90 540
Age (Days) Age ( Days)
Figure 4-15 Effect of total binder content on flexural strength (Air Curing)
76
4-3. COMPARISON OF STRENGTH DEVELOPMENT WITH

PUBLISHED DATA
The results of the present study emphasize the importance of the characteristics of
concrete incorporating high volume fly ash. Compressive strength is in the range 30
to 40 MPa at 28 days, and 40 to 60 MPa was achieved for all the HVF A concretes at
the age of 90 days. In published research, a higher amount of cementitious binder
content, and lower water / binder ratios have been used to produce HVF A concrete
systems. Swamy and Mahmud [59], for example, produced HVF A concrete with
cube compressive strength of23.7 to 63.3 MPa and prism flexural strength of2.6 to
4.5 MPa at 28 days, using a total cementitious binder content of 266 ~ 466 kg/m3 and
water / binder of 0.32 ~ 0.62. Malhotra and Giaccio used [61] cementitious content
of 375 kg/m3 with water / binder 0.31, and the cylinder compressive strength of
HVFA concrete was 32.5 MPa at 28 days, and 41 MPa at 91 days. Langley et al [60]
developed HVF A concrete cylinder compressive strength of 23 to 57 MPa, and 38 to
75 MPa, and 6 to 6.9 MPa, and 8.9 to 9.6 MPa flexural strength at 28 days and 90
days respectively. The HVF A concrete was produced using a total binder content of
225 ~ 400 kg/m3 with water / binder 0.28 ~ 0.49. In Electric Power Research
3
Institute ( EPRI ) research [9], a cementitious content of 365 kg/m with a water /
binder 0.33 was used, and the cylinder compressive strength of HVF A concrete
varied from 27.8 to 42.2 MPa at 28 days, with 50 MPa at 91 days. The prism flexural
strength reached 5.2 ~ 6.5 MPa at 91 days. Table 2-5 in Chapter 2 also shows a
comparison of strength development obtained from other investigators.
4-4. SWELLING OF HIGH VOLUME FLY ASH CONCRETE
Cement paste or concrete cured continuously in water from the time of placing
exhibits a net increase in volume and an increase in mass. This swelling is due to the
absorption of water by the cement gel ; the water molecules act against the cohesive
forces and tend to force the gel particles further apart, with a resultant swelling
77
pressure. In addition, the ingress of water decreases the surface tension of the gel,
and a further small swelling takes place.
Even though swelling associated with ettringite formation is known to be

responsible for both deleterious and beneficial phenomena in portland cement
concretes, this swelling, however, can make the concrete porous and permeable.
In this section a detailed analysis of the swelling of HVF A concretes IS
presented, and the results discussed in terms of the different replacement materials
and the total cementitious content. Under water curing condition, the swelling each
specimen was measured on all four sides, and the results are thus the average of
twelve readings.
Typical swelling test data for all HVF A concretes are shown in Table 4-14
and illustrated in Figs 4-16 and 4-17.
The results show that the final swelling varied over a narrow range, between
100 and 125 micro-strain depending on the total cementitious content. The swelling
initially increased rapidly, after which there was a steady increase with continued
water curing. Between 9 months and one year, there was some sharp increase in
swelling of all the HVF A concretes, beyond which the rate of swelling became more
stabilized till the end of test period. In general, the concrete mix with 450 kg/m3,
showed lower swelling. At the end of 504 days, the swelling value varied from 105 x
10-6 to 125 X 10-6 for the 350 kg/m3 mixes and 100 x 10-6 to 120 X 10-6 for the 450
kg/m3 concretes.
78
Table 4-14 Swelling strain ofHVFA concrete (Water curing)

(Unit: 10-6 m/m )
Age 350F 350SF 350S 450F 450SF 450S

(days)
1 0.00 0.00. 0.00 0.00 0.00. 0.00
7 49.25 38.33 46.25 45.13 36.12 45.23
28 58.33 48.67 56.25 46.45 41.79 52.21
91 68.16 51.75 65.12 53.78 38.52 46.67
189 84.33 56.01 82.33 86.12 44.83 61.25
364 115.21 94.33 113.67 113.21 85.45 97.45
504 124.67 104.67 116.01 118.67 97.33 109.21
The results also show that the HVFA concretes incorporating small amount
of silica fume benefited marginally to suppress swelling. This finding implies that
silica fume acted on the one hand as an intensifier of pozzolanic reaction, and on
other, as an activator of fly ash, while the combined influence of silica fume and fly
ash controlled swelling to an even greater extent.
In view of the swelling of this type at an early age, it can be related to the
hydration process in concrete. According to the swelling theory [67] swelling occurs
because of wetting of the ettringite particles that are of colloidal size ( < 1 ~m in
length ), from a topochemical growth of ettringite pushing against solid material in
the concrete. The early hydration reaction is located at the acceleration stage of the
calorimetric curve which is due to the formation of ettringite as the hydration of C3A
and sulphate ion occurs, and the extra pressure of the inner part in C3 S grains
accelerates the concentration of calcium, in which a large amount of CH crystal as
well as C-S-H gel will form.
79
1~ r----------------------------------------------,
120
-
\0
I
90
.~
.j:j
00
.---~
Q
( l)
60
~
00
30
350F 3508F 3508
)( 0 0
o~~~~~--~~~~~--~~~~~~--~~~~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-16 Variation of swelling of HVFA concrete-350 kg/m3 mixes
150 .----------------------------------------------,
120
-
\0
I
Q 90
·ca
.j:j
00
.---~
Q
( l)
60
~
00
30
450F 4508F 4508
•
o~~-L~~L-~~-L~~--L-~~-L~--~~~~~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
3
Figure 4-17 Variation of swelling of HVFA concrete-450 kg/m mixes
80
A companson of the influence of the total cementitious content on the

swelling ofHVFA concretes under water curing is shown in Figs 4-18, 4-19, and 4-
20. These results show that the binder content has some remarkable effect on the
3
development of swelling. In general, 450 kg/m mixes had a lower swelling than
3
those made with 350 kg/m mix. The difference in swelling between the two types of
concretes is probably due to the higher superplasticizer content used in the 450 kg/m3
3
mixes. In the present study, the 450 kg/m concrete mixes were too glutinous to mix
and cast. Hence, a relatively high superplasticizer dosage was needed for these mixes.
With a higher superplasticizer dosage, the ettringite may have grown into the water-
filled space, and resulted in reduced swelling strains [67].
150
350F 450F
x
120
\0
I
0
....-4
0 90
'ed
~
00
.-..........
OJ:)
0
Q) 60
~
00
30
o~~-L~~~ __ L-~~~~~--L-L-~~~~~--~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-18 Effect of total binder content on swelling-350F/450F
81
Chapter 4 Test Results and Discussion of Part J 82
150
350SF 450SF
e
120
\0
I
0
,.......;
~
90
·a /~
~
r/J
bI)
~
......
.~
...... 60
(I)
~
r/J
30
o~~~~~--~~~~~--~~~~~~--~~~~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-19 Effect of total binder content on swelling-350SF/450SF
150
350S 450S
e •
120
\0
I
0
,.......;
90
.S
t\l
~
r/J
bI)
~
......
.~
..- 60
(I)
~
r/J
30
o~~-L~~ __ ~~~-L~--L-~J-~~~--L-~-L~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-20 Effect of total binder content on swelling-350S/450S
82
4-5. DRYING SHRINKAGE OF HIGH VOLUME FLY ASH

CONCRETE
Drying shrinkage of concrete is caused by the contraction of the calcium silicate gel
in the hardened paste as the moisture content is reduced. As concrete dries, the water
lost first comes from the capillaries and causes little shrinkage. After the capillary
water has been lost, the water is removed from the cement gel pores and noticeable
drying shrinkage occurs. While many factors determine the magnitude of drying
shrinkage, the materials that mostly affect drying shrinkage are water, the
cementitious paste, admixtures and aggregates [68].
In this section, the results of the tests on drying shrinkage of HVF A concretes
are presented. The drying shrinkage specimens were initially moist cured for 7 days,
after which they were transferred to the ambient environment in the laboratory and
monitored over a period of 504 days at regular intervals of 7 days. The effect of
variations in different replacement methods and total cementitious content was also
investigated. The experimental drying shrinkage data are evaluated with the aid of
Ross' formula [13]. The relationship between the drying shrinkage and the swelling is
also reported.
4-5-1. Effect of different mixes on drying shrinkage
Drying shrinkage values are expressed as linear strain as a function of age and shown
in Table 4-15 and illustrated in Figs 4-21 and 4-22.
At early ages, the HVF A concretes have not fully matured, and the high early
drying shrinkage strain obtained at this stage were expected. A high proportion of the
drying shrinkage strain for all the air dried HVF A concretes occurred during the
initial 28 days, after which the shrinkage gradually increased with increase in age and
then became somewhat stabilized with time until the end of the test period.
83
Compared with all HVF A concretes, it can be seen that the drying shrinkage
strain results obtained in this study are relatively high. At the end of the test period,
the drying shrinkage strain of all the HFVA mixtures varied between 550 and 640
microstrains. According to the data obtained in this investigation, mixtures with
cement replaced partially by fly ash alone, i. e. mixes, 350F, 450F showed greater
resistance to drying shrinkage. Mixes 350S and 450S with large amounts of fly ash in
the concretes showed slightly higher drying shrinkage strain, whereas mixes 350SF
and 450SF showed the highest drying shrinkage strain values at the age of 504 days.
The addition of silica fume did not thus seem to restrain significantly the drying
shrinkage of air dried HVFA concretes. In contrast, the swelling of HVF A concretes
stored in water indicated that concretes incorporating of small amount silica fume,
350SF and 450SF, showed a significant reduction in swelling, and had good ability to
suppress swelling.
The reduction in swelling could also be attributed to the lower water / binder
ratio of the HFVA concrete, and the changed pore size distribution of the matrix,
thereby slowing the rate of swelling. Normally, swelling and drying shrinkage are
related - the higher the swelling reduction, the greater its the drying shrinkage [69].
Table 4-15 drying shrinkage strain ofHVFA concrete (air curing)

(Unit :10-6 mlm )
Age 350F 350SF 350S 450F 450SF 450S

_( d<!)'s )
1 0.00 0.00 0.00 0.00 0.00 0.00
7 50.67 41.12 47.12 29.67 38.33 46.33
28 -176.70 -366.15 -340.15 -303.12 -362.45 -394.33
91 -387.33 -511.33 -450.67 -481.15 -533.67 -549.67
189 -475.33 -596.12 -557.15 -554.12 -616.12 -587.33
364 -530.21 -605.21 -565.12 -580.33 -615.67 -600.33
504 -550.21 -636.33 -590.12 -575.33 -642.33 -612.67

Chapter 4 Test Results and Discussion of Part 1
200 r-------------------------------------------~
350F 3508F 3508

100 )( o o
O~--------------------------------~
'D
I
o
-" (100)
~
j (200)
00 (300)
I:lI)
.S
g (400)
(500)
(600)
(700) L...-..L---'---1..---L-.J._L_..L---'------'---L----L'--'--..L--l..----L----L..----L-----1_ _.l-.....l
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-21 Variation of drying shrinkage ofHVFA concrete-350 kglm3 mixes
200 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
450F 450SF 450S

100
•
O~-------------------------------~
'D
I
o
-" (100)
Q)
~ (200)
]
00 (300)
I:lI)
.S
g (400)
(500)
(600)
(700) L-...L.---L..--L..~___'__L-....L._-1.---L.---'-___'__ _'___...L.._....L_____'___L...___'__ _L_...l....__...J
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
3
Figure 4-22 Variation of drying shrinkage ofHVFA concrete-450 kglm mixes
85
Figs 4-23 and 4-24 represent the swelling / shrinkage characteristics for
HVF A concrete from the present study. It can be seen from the figures, that the
higher the swelling of 350F and 450F, the lower is its drying shrinkage strain. To
investigate the relationship between drying shrinkage and swelling of HVF A
concretes, a plot of drying shrinkage data versus swelling at approximately the age of
504 days is presented in Fig 4-25. Regression analysis of the data shown in the figure
produced the following relationship :
y = 879.69 - 2.59 x (r= - 0.98 ) ---------- ( 4-9 )
where y is the drying shrinkage of HVFA concretes in microstrain, x is the swelling

ofHVF A concretes in microstrain, and r is the correlation factor.
The figure indicates that the good inverse correlation between drying
shrinkage and swelling is linear.
As was mentioned in the preceding review on drying shrinkage, a careful

study of the pore size distribution may be useful in explaining some aspects of drying
shrinkage which are not always well understood. There is reason to believe that
drying shrinkage is the result of processes taking place with the finer pores of the
system, i.e. drying of water from the mesopores causing capillary effects and from the
micropores inducing drying shrinkage strains which may be the result of mechanisms
such as changes in disjoining pressure or interlayer space [70].
Table 4-16 shows typical data from the present study of M. I. P. test on all
HVF A concrete air dried specimens. It can be seen from the table, that at the age of
18 months, the percentage gain in coarse pore size of concretes with cement replaced
partially by fly ash alone, 350F and 450F, is higher in comparison with those
incorporating both fly ash and silica fume concretes. The drying shrinkage behaviour
of HVF A concretes in the present study is explained in terms of coarse and fine pore
size with increase in coarse pore size resulting in a decrease in drying shrinkage and
increase in fine pores producing increased drying shrinkage. The results on drying
shrinkage thus confirm those observed in published literature [71].
86
200
100 ®
'9 ~§ <>
0 0
.-4
d'
.~ (100)
350F
en
)(
~ (200)
350SF
o
] (300)
-.S
en
.-.
.-.
(400)
X
350S
o
Q) 0
~ (500) <>
(600)
(700)
0 112 224 336 448 560
Age (Days)
Figure 4-23 Swelling / drying shrinkage characteristics-350 kglm3 mixes
200 ~--------------------------------------------~
t
\0
I
100
:.
o o ~--------------------------------------------~
.-4
.~~ (100) +
450F
en
Q)
~ (200)
450SF
~
-.S
] (300)
en 450S
•
(400)
.-.
.-.
Q)
~ (500)
+
(600) •
(700) L....._--'-_---1.._ _. L . . . . - _ - L - _ - - L_ _L - - _ - - ' - _ - - - L . ._ _L..==-...J
o 112 224 336 448 560

Age (Days)
3
Figure 4-24 Swelling / drying shrinkage characteristics-450 kg/m mixes
87
6W r-----------------------------------______~
640
-
1.0
I
o
o
o
00
y = 879.686-2.5860 x
r= - 0.9799
560
520 ~~--~--~--~--~~--~ ___ L_ _~_ _L_~_ _~

80 90 100 110 J?O 130 140
Swelling Strain, 10
Figure 4-25 Relationship between drying shrinkage and swelling
Table 4-16 Typical % of meso pore of high volume fly ash concrete under air
curing
(Unit: % )
Pore size > 0.05 j..lm 0.05 j..lm ~ 0.0025 j..lm
Mix 28days 90days 18months 28days 90days 18months
350F 61.16 65.40 78.03 38.84 34.62 21.97
350SF 62.11 43.78 67.85 37.89 56.22 32.15
350S 37.93 31.26 61.21 62.07 68.74 38.79
450F 51.45 42.38 71.87 48.55 57.62 28.13
450SF 62.05 60.36 67.69 37.95 39.64 32.31
450S 47.60 36.41 61.03 52.40 63.59 38.97
88
Although the HVF A concrete exhibited high drying shrinkage strain values
early in its life, these drying shrinkage strain levels are either comparable to or in a
manner similar to those observed by Malhotra and his co-worker [9,43]. They
obtained a range of drying shrinkage from 400 ~ 600 microstrains at the end of 448
days with similar concretes under essentialy the same test conditions.
The effect of total binder content of all HVF A concretes on drying shrinkage
IS compared in Figs 4-26, 4-27, and 4-28. The figures show that the total
cementitious content has only a marginal influence on the values of drying shrinkage.
As expected, the 450 kg/m3 mixes had a slightly higher drying shrinkage strain than
with 350 kg/m3 . The higher drying shrinkage of 450 kg/m3 mixes is probably due to a
combination of two factors - a higher binder content and superplasticizer dosage
[42]. In addition, at a constant water I binder ratio, drying shrinkage increases with
an increase in the cementitious content because this results in a large volume of
hydrated cementitious paste which is liable to drying shrinkage [69].
200 ~------------------------------------------~
350F 450F
100 )(
O~-------------------------------------------i
\0
I
o
.......~ (100)
Q)
bO
~ (200)
]
00 (300)
00
a
.S
(400)
(500)
(600)
(700) L-..l.----L--L--L.--l...----1--''---.L-..-J.--L--L----'-----'----'---L...-.1.--.-'----'---'----'
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-26 Effect of total binder content on drying shrinkage-350F/450F
89
200 r-------------------------------------------~
350SF 450SF
100 o
O~-------------------------------------------~
o
......" (100)
(!)
E~
(200)
]
r:Fl (300)
gf
a
......
(400)
(500)
(600)
(700) L...--'---'----'----'-----l_ _L..--'---'----'----'-----l_L..-...L--1----L----L..---L.L-.L-..L.-..J
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-27 Effect of total binder content on drying shrinkage-350SF/450SF
200 r--------------------------------------~
350S 450S
100 o •
o~------------------------------------------~
o
......" (100)
(!)
bO
~ (200)
~
.\::
..e
r:Fl (300)
gf
a
......
(400)
(500)
(600)
(700) L-....L.---L-.---'----L--'_ _L-....L.---L-.---'----L----l--.L...-....L.---L---'-----L----l--L..-..l....-..-I
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-28 Effect of total binder content on drying shrinkage-350S/450S
90
4-5-2. Prediction of drying shrinkage on HVFA concrete
Many kinds of equations including ACI, CEB formulae, Meyer, and Bazant have
been developed to predict the drying shrinkage of conventional concrete. In the
analysis of the drying shrinkage of the HVF A concrete under investigation, a
hyperbolic expression suggested by Ross has been used [69].
Ross's equation can be expressed as :
t
c = -------------------- ( 4-10 )
a+bt
where c is the drying shrinkage of HVFA concrete in microstrain, t is the time after
initial readings for drying shrinkage in days, and a and b are constants.
when t = 00, then c = 1 / b, i. e. 1 / b is the limiting value of drying shrinkage.

The symbols a and b represent constants determined from experimental results. A
plot of test results of t / c against "t" gives a straight line relation of slope "b" and an
intercept "a".
Figs 4-29 and 4-30 show the curves for determining the constants "a" and "b"
for all HVF A concretes undergoing drying shrinkage. Table 4-17 shows the constants
derived from these results. In the same table, typical results of the measured and
predicted values of drying shrinkage at various ages for all HVF A concretes are also
shown.
From the results, during the first 3 months, the drying shrinkage strains
obtained through Ross's formula are substantially higher compared to experimental
results. However, as can be seen from the table, there is good agreement between the
measured and predicted values, especially for ages greater than 6 months. The drying
shrinkage strains measured on all HVF A concretes are comparable to predicted
values, i. e., about 95 % of the average percentage of the latter.
91
350F 3508F 3508

)( o o
0.8
(])
~
~ 0.6
]
00 /
--
~ 0.4
<
o ~~~~~--~~~~--~~~~--~~~~~--~~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-29 The curve for determining the constants a and b

( 350 kglm3 mixes)
450F 4508F 4508

• +
0.8
--
~ 0.4
<
0.2
o L-~-L~ __L-~-L~~L-~~-L~--~~~~--~~~~
o 56 112 168 224 280 336 392 448 504 560
Age (Days)
Figure 4-30 The curve for determining the constants a and b

( 450 kglm 3 mixes)
92
Table 4-17 Comparison between measured and predicted drying shrinkage

( Unit :10-6 m/m )
A. 350 kg/m3 mixes

Mix 350F 350SF 350S
3 3
a=99.73 x 10- a = 41.51 x 10- a = 44.20 x 10-3
3
b = 1.58 X 10- b = 1.50 X 10-3 b = 1.60 X 10-3
Age Measured Predicted Measured Predicted Measured Predicted
( days)
14 69.67 114.90 156.12 223.96 178.67 210.20
28 176.70 194.48 366.15 335.29 340.15 314.61
91 387.33 373.70 511.33 511.21 450.67 479.95
189 475.33 474.46 596.12 581.52 557.15 545.30
364 530.21 539.38 605.12 619.56 565.12 580.92
504 550.21 562.47 636.33 631.97 590.12 592.52
B. 450 kg/m3 mixes

Mix 450F 450SF 450S
3 3
a = 44.82 x 10- a = 33.99 x 10- a = 28.29 x 10-3
b = 1.62 X 10-3 b = 1.50 X 10-3 b = 1.57 X 10-3
Age Measured Predicted Measured Predicted Measured Predicted
( days)
14 125.25 207.41 193.33 254.59 200.12 278.50
28 303.12 310.49 362.45 368.47 394.33 387.54
91 481.15 473.37 533.67 533.76 549.67 531.67
189 554.12 538.46 616.12 595.29 587.33 581.50
364 580.33 573.68 615.67 627.60 600.33 606.90
504 575.33 585.16 642.33 637.98 612.67 614.96

Remarks:
Predicted drying shrinkage = age
a+b x age
4-5-3. Comparison of drying shrinkage with published data
Table 4-18 shows a comparison of drying shrinkage from the present investigation
with the results of other researchers [9,29,44,59,72]. Data from the table show that
93
Chapter 4 Test Results and Discussion of Part J 94
there is a wide range of data on drying shrinkage reported on fly ash concrete mainly
due to the different concretes tested, differences in type of curing, and age of testing.
When all the fly ash concretes under various curing regimes are considered, the
drying shrinkage varies from 385 to 642 microstrain. From these data, it is obvious
that it is impossible to make in any general conclusions covering all types of fly ash
concretes. It is worth to point out, however, that the results of drying shrinkage
obtained in the present study are similar to those reported by EPRI and Malhotra ;
they are, however, much higher than the results reported by Swamy. The reason for
this difference is probably due to the mix proportions and the water / binder ratio.
Table 4-18 Comparison of drying shrinkage with published data
Researchers Fly ash W/C Cube St. Curing Age at last Drying
content (28d) reading shrinkage
(% ) MPa conditions (days) [( microstrain )
0046
Cripwell 35 44.5 20°C + 70 % R. H 300 485
[29] 0.35
" 53.0 " " 400
0043 26.8* 23°C + 70 % R. H 91 500
Ward 40
0041 26.0*
[72] 50 " " 580
0.32 35.0* 7 wet + air curing 448 500
Malhotra [44] 58
0.32~0.62
Swamy 50 24~63 Outside 550 385 ~ 422
[59] "
II
" Laboratory 730 460 ~ 485

EPRI 58 0.33 27.8* 7 wet + air curing 448 454
[9] "
II
32.5* " " 606
" " 38.5* " " 457
" " 28.8* " " 433
"
II
38.1 * " " 504
"
II
41.4* " " 535

Author 350F( 57%) 0040 28.4 7 wet + air curing 504 550
350SF( 51%
II
29.9 " " 636

350S( 97%) " 38.9 " " 590
450F( 56%) " 28.6 " " 575
450SF( 51% " 38.1 " " 642
450S( 97%) " 39.3 " " 613
Remarks : * - Cylinder strength
94
4-6. DYNAMIC MODULUS OF ELASTICITY OF HVFA

CONCRETE
Modulus of elasticity is important in structural design of reinforced concrete, and is

used primarily to evaluate soundness of concrete as a measure of durability. It is
often used as a non-destructive test to determine the progression of internal changes
in the material, as for example, a consequence of repeated freeze-thaw cycles, and
ageing, or of chemical attack.
For this purpose, modulus of elasticity measurements are very convenient,

since they are non-destructive, and since there is also a general ( if empirical )
relationship between the modulus of elasticity and compressive strength. Like many
other structural materials, concrete is, to a certain degree, elastic. The modulus of
concrete represents its response to elastic deformation, and it also exhibits a very
high sensitivity in reflecting the internal changes in the structure of concrete.
There is now absolutely no doubt that the use of high strength and high
performance concrete in construction will continue to grow. The elastic deformation
of high strength / high performance concrete is of particular importance. Because the
modulus of elasticity of the strong hardened cement paste and of the aggregate differ
less from one another than in medium strength concrete, the behaviour of high
strength / high performance concrete is more monolithic, and the strength of the
aggregate-matrix interface is higher.
In this section the main results deal with the characteristics of dynamic
modulus of elasticity in different mixes and different binder contents concrete under
both types of curing conditions. For purpose of comparison the results of the normal
ope concretes are also discussed. A comparison of the correlation between the
dynamic modulus of elasticity and compressive strength of HVF A concrete is also
reported.
95
4-6-1. Effect of curing condition and different mixes on dynamic

modulus of elasticity
Under both curing conditions at various ages, the dynamic modulus of elasticity of
HVF A concretes with both cementitious contents is illustrated in Figs 4-31, 4-32, 4-
33, and 4-34. Tables 4-19 and 4-20 summarize typical data of all HVFA concretes
under water and air curing, respectively.
Table 4-19 Dynamic modulus elasticity ofHVFA concrete (Water curing)

(Unit: GPa)
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 23.01 23.11 24.25 38.11 32.00 31.01
7 35.25 35.01 34.21 45.07 46.01 45.78
28 36.03 38.85 39.69 46.07 50.25 51.21
91 39.50 42.29 41.96 50.78 54.83 54.52
189 44.49 43.85 43.16 55.04 56.96 55.97
364 45.70 45.40 44.59 58.85 59.24 58.01
504 46.45 46.22 45.18 59.51 60.14 58.75
Table 4-20 Dynamic modulus elasticity ofHVFA concrete (Air curing)

(Unit: GPa)
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 22.69 22.43 23.99 37.92 31.15 30.85
7 35.04 34.03 34.99 44.85 45.57 45.56
28 35.50 36.58 37.86 45.37 47.96 47.42
91 34.30 35.60 36.44 43.24 46.37 45.45
189 33.60 34.65 35.76 42.09 45.61 44.68
364 33.55 34.72 36.08 42.35 45.62 44.95
504 34.20 35.16 36.30 43.53 45.91 45.68
96
The results indicate that the modulus increased rapidly during the first few
days, it then increased slightly with further curing, and finally reached a stable value
with continuing hydration reaction. Data from the tables show that the modulus of
elasticity of moist cured specimens is consistently higher than that of the dry cured
specimens. The results also show that the total binder content is a predominant factor
in determining, the value of dynamic modulus. Concrete with 450 kglm3 was superior
3
to those with 350 kglm under both types of curing. Irrespective of the cement
replacement methods, the difference between 350 kglm3 and 450 kglm3 mixes under
both curing conditions is seen to be identical. At the age of 504 days, the difference
in modulus of elasticity values between 350 kglm3 and 450 kglm3 mixes is about 20 -
25 % for both curing conditions. The total binder content generally controls the
concrete making properties of the aggregate, because it determines the amount of
binder paste needed to cover the aggregate surface. A high binder content of
concrete means that it can be significantly enhanced via the physical mechanisms of
improved packing of hydration products, so that the strength of the matrix is high.
Because hydration starts at the surface of the binder particles, it is the total surface
area of binder that represents the material available for hydration. Therefore, the
strength between the coarse aggregate and the matrix, and the elastic properties of
aggregate have considerable influence on the dynamic modulus of elasticity of
concrete.
The densifYing effect of fly ash and silica fume are demonstrated in the data
presented here. However, Mix-450SF concrete with silica fume had a higher dynamic
modulus of elasticity, whereas Mix-450S with a large amount of fly ash had a slightly
lower modulus than the concrete made of 450F.
97
Chapter 4 Test Results and Discussiun of Part I 98
70 ,--------------------------------------------
20
350F 350SF 350S 450F 450SF 450S
)( o o •
10 ~~~~~~~ __ ~~~_L~~~~L_~~_L_L~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-31 Variation of dynamic modulus ofHVFA concrete (Water curing)
70 ~------------------------------------------~
60
20 450S
350F 350SF 350S 450F 450SF
)( o o •
10 L-~L_~~~~~_L~_L~~~~~~~~~~
o 56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-32 Variation of dynamic modulus of HVFA concrete ( Air curing)
98
70 70
---ca 60 ca
6 V- @) 60
'-" '-"
fJl 50 l------.-.-.--.-..----.-- fJl 50
.§ .§
"d 40 "d 40
~ ~
.~ 30 -~ 30
~20 [ 3:~45OC [ S20
Cl
10 10
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
,-... 70 ,.-... 70
<:d ca
6 60 6 60
'-" '-"
~
;3
~ 40
50
- 4050
~
;::1
"d
Alli1~~f3CkJC 8-60 0 0 0 0 0 0 0 O-£)
r o
20
10
0 112 224
Days
336 448 560
r o
20
10
0 112 224
Days
336 448 560
Figure 4-33 Effect of total binder content on dynamic modulus ( Water curing)
,-...
70 ,-...
70
<:d <:d
6 60 @) 60
'-" '-"
50 .-..__ --..__..... _....-- ...__.....__ ....-.
fJl fJl 50
.§ .§
"g 40 "d 40
;:;s ~
r o
20
10
0 112
[ 350C 450C [
224
Days
336 448 560
-~ 30
S20
10
0 112 224
Days
336 448 560
~
70 ,.-...
70
<:d ca
6 60 6 60
'-'
~ 50
"d 40
-
f ........................ • • • •
'-'
.a 40
fJl
't:I
50 ,... - --""
l~
~ I~",
~
.L'. ~~~
r
~
A/,-/\..6.+'>
!? ~
o
-I 30
5
20 !> I 35!F45!F I 20 13~S4!S I
10 10
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
Figure 4-34 Effect of total binder content on dynamic modulus ( Air curing)
99
From the Table 4-19, the dynamic modulus of elasticity of HVF A concretes
ranged from 36 to 51 GPa at 28 days, from 39 to 55 GPa at 91 days, and from 45 to
60 GPa at 18 months for both cementitious contents. Considering the levels of
compressive strength of HVF A concrete at these three ages, the dynamic modulus of
elasticity values are rather high. As mentioned in the previous section, the generally
high dynamic modulus of elasticity values are due to the unhydrated fly ash particles
acting as fine aggregate. As for the chemical reaction of silica fume, because of high
surface area and high content of amorphous silica in silica fume, this highly active
pozzolan reacts more quickly than ordinary pozzolans [73].
From the Fig 4-32, air dried specimens showed slight retrogression of
dynamic modulus of elasticity and show similar trends to changes in strength when
concretes are exposed to a drying environment. As in the strength development
studies, the effect of curing conditions on dynamic modulus of elasticity is significant.
At the end of 504 days, the dynamic modulus of elasticity of HVF A concretes under
air drying ranged from 34 to 46 GPa compared to 45 to 60 GPa for water cured
specimens. Concretes made with both fly ash and silica fume exhibited slightly higher
values than concrete made of350F and 450F.
The relative rates of development of dynamic modulus of HVFA concrete

cured in water and air are illustrated in Table 4-21.
Table 4-21 Influence of air curing on dynamic modulus of elasticity

( Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

( days)
14 100 99 100 98 99 98
28 98 94 95 98 95 93
91 87 84 87 85 85 83
189 75 79 83 76 80 80
364 73 76 81 72 77 77
504 74 76 80 73 76 78
Remarks : Water cured specimen is 100 %
100
Chapter 4 Test Results and Discussion o/Part 1 101
The percentage reduction in dynamic modulus is about 5 % of average values

for all HVF A concretes when compared to the moist cured concrete at the age of 28
days. Even though air curing showed a little retrogression of dynamic modulus, a
modest increase of modulus with continued air drying was also observed. At the end
of air curing, the average percentage of the dynamic modulus of elasticity of all
HVF A concretes is about 76 % of that under water curing.
It would then be interesting to compare the dynamic modulus of elasticity

development of this HVFA concrete with that of normal OPC concrete with age and
under both curing conditions. Table 4-22 represents some typical data of dynamic
modulus of elasticity for two normal OPC concretes under both types of curing
conditions. The rate of gain of modulus as a % of 28 day modulus for HVFA
concretes under both curing conditions is shown in Tables 4-23 and 4-24.
Table 4-22 Dynamic modulus elasticity of normal ope concrete
Mix 350 kg/m3 mixes 450 kg/m3 mixes

GPa Dynamic % of28 day Dynamic % of28 day
or modulus development modulus development
% (GPa) (%) (GPa) (% )
{da)'s) curing curing curing curing curing curinZ curing curing
1 42.77 42.82 92 93 54.77 55.01 92 94
7 44.84 44.95 96 97 57.94 58.04 97 99
28 46.52 46.23 100 100 59.54 58.82 100 100
91 47.48 46.13 102 99 62.30 58.68 105 99
189 48.35 45.98 104 99 62.11 58.26 104 99
364 49.02 45.78 105 99 63.34 57.71 106 98
101
Chapter 4 Test Results and Discussion o[PartJ 102
Table 4-23 Rate of gain of modulus as a % of 28 day modulus ( Water curing)

(Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 64 59 61 82 64 61
7 98 90 86 98 92 89
28 100 100 100 100 100 100
91 110 109 106 110 109 106
189 125 113 109 119 113 109
364 127 117 122 128 118 113
504 129 119 114 129 120 115
Table 4-24 Rate of gain of modulus as a % of 28 day modulus ( Air curing)

(Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 64 61 63 84 65 65
7 99 93 92 99 95 96
28 100 100 100 100 100 100
91 97 97 96 95 97 96
189 95 95 94 93 95 94
364 95 95 95 93 95 95
504 96 96 96 96 96 96
From the results, it can be seen that, under continuous moist cunng
conditions, the percentage increase of dynamic modulus of elasticity of normal OPC
concrete from 28 day to one year is 5 % to 6 %. The corresponding average increase
for the HVFA concrete was 13 % to 28 %. The percentage increment of dynamic
modulus of elasticity of HVF A is thus about 3 - 4 times that of the normal OPC
concrete. As water curing continued, the HVF A concrete appeared to undergo some
102
more hydration reaction than the ope concrete although equivalent mechanical
properties were never reached.
At the end of 504 days, the average percentage loss on dynamic modulus for
dry cured specimens of HVF A concrete is quite low, about 4 %, after the 28 day
value. The decrease in modulus of HVF A concrete is thus very small and almost
negligible. It is worth recalling the effect of prolonged air drying on strength
development. With a short period of seven days moist curing, it is possible not only
make the full achievement of the strength potential of HVF A concrete, but in effect
also preserve the complete development of its modulus of elasticity [6].
4-6-3. Relationship between the dynamic modulus and compressive

strength
The modulus elasticity is greatly influenced by the strength of concrete, stifthess of

the aggregate, and by the volumetric proportion of aggregate in the concrete. It is
also influenced by curing conditions, the age of the concrete, and mix proportions of
the concrete. Although individual studies have shown that the modulus elasticity
increases with an increase in the compressive strength of concrete, there is no
agreement on the precise form of the relationship [74,75,76].
When high volume fly ash / high strength concrete began to be developed,
efforts were made to see whether the existing relationships could be used to predict
high volume fly ash concrete modulus of elasticity or new formulae had to be
developed.
The relationship between the dynamic modulus of elasticity, ED, and

compressive strength, fe, is generally by the following equation :
ED = a feb ------------- ( 4-11 )
where ED is the dynamic modulus of elasticity of concrete III GPa, fe IS the

compressive strength, in MPa, and a and b are constants.
103
The relationship between dynamic modulus of elasticity and compressive

strength for moist and dry cured specimens, at ages of 1 day to 18 months, is shown
in Figs 4-35 and 4-36. Due to the influence of the total cementitious content on
concrete and the wet specimens having higher dynamic modulus of elasticity than the
dried specimens, it is recommended that separate equations be used for the two
conditions.
Regression analysis of the data shown in these figures produced the following
relationships :
Under water curing
3
ED = 14.07 fc°.3O (350 kg/m mixes) (r = 0.98) --------- ( 4-12 )
30
ED = 21.50 fco. (450 kg/m3 mixes) (r = 0.96) --------- (4-13 )
Under air curing
ED = 17.94 fco. 18 ( 350 kg/m3 mixes) (r = 0.88 ) --------- ( 4-14 )
ED = 28.32 fcO. 13 (450 kg/m3 mixes) (r = 0.79) --------- (4-15)
Irrespective of the total cementitious content, for the wet cured concrete, the
dynamic modulus of elasticity, ED, increases approximately with the cube root of its
curing compressive strength, fc . It is clear that dynamic modulus of elasticity can be
accurately predicted from compressive strength, for moist cured specimens. Each
individual equation produces a high correlation factor between the two parameters
under water curing and there is a significant correlation at the 99 % confidence level
between dynamic modulus of elasticity and compressive strength of HVFA concrete.
In Fig 4-36, the relationship is observed to be markedly influenced by both the

total cementitious content and the air drying. Although these correlation coefficients
are somewhat low, there is a significant correlation at the 96 % confidence level
between dynamic modulus of elasticity and compressive strength of HVF A concretes
under air curing.
104
Chapter 4 Test Results and Discussion of Part I
70 ,------------------------------------------
3
350 kg / m 450 kg / m3
o •
60
r:o:s
fj
~' 50 • •
'"S
"0
0 ••• o
~
u •• o
.~ 40
~ • o
6' o
o
o 0
30
20 ~~~--~~--~~--~~~--~~--~~--~~~
o 10 20 30 40 50 60 70 80
Compressi ve Strength, MPa
Figure 4-35 Relationship between dynamic modulus elasticity and compressive

strength ( Water curing)
70 .---------------------------------------------~
3 3
350 kg / m 450 kg / m
o •
60
•• •
. ... ••
• •••
30
20 LL~_L _ _~~~_ _L_~~~_ _~~~_ _~~~~
o 10 20 30 40 50 60 70 80
Figure 4-36 Relationship between dynamic modulus elasticity and compressive

strength ( Air curing)
105
Chapter 4 Test Results and Discussion o{Part I 106
On the basis of the above results, the relationships between dynamic modulus
of elasticity and compressive strength are curvilinear since the dynamic modulus of
elasticity is more dependent on the moisture gradient.
4-6-4. Comparison of modulus elasticity prediction
While many results of tests for modulus of elasticity of concrete have been published,
there are only a few where the dynamic modulus of elasticity on HVF A concrete have
been tested and its relationship reported. Most of the researchers [9,27,43,59] have
presented the results of static modulus of elasticity. For this reason, Eqn. 4-16, as
suggested by the British Code for design of concrete structures, CP 110 : 1972 is
also used in the present investigation. The equation may be more convenient to
estimate the static modulus of elasticity, Ec, from the measured dynamic modulus of
elasticity, ED.
Ec = 1.25 ED - 19 (Both moduli being expressed in GPa) --------- ( 4-16 )
A summary of the functions proposed by various researchers including those

obtained from the results for all the HVFA concretes are shown in Tables 4-25 and 4-
26, and compared in Fig 4-37. Comparison is also made with the relationship
predicted by ACI 318-89 and CP 110, as follows:
Ec = 4.73 fc 0.50 ( ACI 318-89 ) ----- ( 4-17 )
Ec = 9.10 fco. 33 ( CP 110 ) ----------- ( 4-18 )
Form the figure, good agreement was found between the predicted and
3
measured values for the modulus of elasticity ofHVFA concrete made of350 kg/m
mixes, but, it was lower than the reported by EPRI [9] and Swamy [59]. However,
the levels of compressive strength of the HVF A concrete at 28 and 504 days, and the
modulus of elasticity of 450 kg/m3 mixes are rather high. At the end of 18 months,
the test results are higher than the ACI Building Code 318-89 calculated values by
about 43 ~ 49 % for the 450 kg/m mixes.
3
106
Table 4-25 Comparison of predicting modulus elasticity and published data

under water curing ( 28 days )
Measured values of compressive strength and modulus b Author

Mix 350F 350SF 350S 450F 450SF 450S
lL"
e (MPa) 33.2 37.1 41.4 31.0 39.1 39.5
*Ee= 1.25 ED - 19 (GPa) 26.3 29.6 30.6 38.6 43.8 45.0
Calculated value ( GPa )
lEe = 4.73 feo. s (ACI 318-89) 27.3 28.8 30.4 26.3 29.6 29.7
lEe = 9.10 fe°.33 (CP 110) 29.2 30.4 31.5 28.6 30.9 31.0
iEe= 10.95 feo. 3 (Swamy) 31.4 32.4 33.5 30.7 32.9 33.0
lEe = 8.79 feO. 39 (EPRI ) 34.5 36.0 37.6 33.5 36.7 36.9
( *Ee - Ee ) / Ee (0/0)
ACI318-89 -4 +3 +1 +47 +48 +52

CP 110 -10 -3 -3 +35 +42 +45
Swamy [59] -16 -9 -9 +26 +33 +36
IEPRI [9] -24 -18 -19 +15 +19 +22
Table 4-26 Comparison of predicting modulus elasticity and published data

under water curing ( 504 days )
Measured values of compressive stren ~h and modulus bv Author

Mix 350F 350SF 350S 450F 450SF 450S
~e (MPa) 62.2 60.1 64.8 62.4 63.8 67.4
*Ee= 1.25 E D - 19 ( GPa) 39.1 38.8 37.5 55.4 56.2 55.4
Calculated value ( GPa )

lEe = 4.73 feo. s ( ACI 318-89 ) 37.3 36.7 38.1 37.4 37.8 38.8
lEe = 9.10 feo. 33 ( CP 110 ) 36.1 35.6 36.6 36.1 36.4 37.0
lEe = 10.95 feO. 3 (Swamy) 37.9 37.5 38.3 37.9 38.1 38.8
lEe = 8.79 fe°.39 (EPRI ) 44.0 43.4 44.7 44.1 44.5 45.4
( *Ee - Ee ) / Be (%)
~CI 318-89 +5 +6 -2 +48 +49 +43
CP 110 +8 +9 +3 +53 +54 +50
Swamy [59] +3 +4 -2 +46 +48 +43
EPRI [9] -11 -11 -16 +26 +26 +22
107
70 r---------------------------------------
Author Author
ACI318-89 CP 110 Swamy EPRI
350 kg/n? 450 kg/of
60 • o o •
Author
•
Swamy
20 CP 110
10 ~~~ __ ~~ __ ~~ __ ~~~ __ ~~ __ L_~~L_~_ _
o 10 20 30 40 50 60 70 80
Figure 4-37 Comparison of predicting modulus elasticity and published data
4-7. ULTRASONIC PULSE VELOCITY OF HVFA CONCRETE
The ultrasonic pulse velocity through concrete is the outcome of the time taken by
the pulse to travel through the hardened cement paste and through the aggregate.
The pulse velocity change is useful information for understanding the development of
the inner structure of the hardened concrete. The test is , therefore, useful to detect
cracking, voids, deterioration due to frost or fire, and the uniformity of concrete in
similar elements.
The process of hydration can provide strength because of microstructure

growth. The ultrasonic travel time then gradually decreases since cement paste
gradually transfers from a plastic to a solid, and that accelerates the ultrasonic pulse
velocity.
In this section a detailed analysis of the ultrasonic pulse velocity on HVF A

concretes will be discussed, and the effect of curing conditions and different mixes on
108
their values will be described. The experimental ultrasonic pulse velocity data on
HVF A concretes and normal ope concrete are also compared. A relationship
between the ultrasonic pulse velocity and the dynamic modulus of elasticity is also
described.
4-7-1. Effect of curing condition and different mixes on HVFA

concrete
The variation of ultrasonic pulse velocity of HVFA concretes with both binder
contents under the two curing regimes is illustrated in Figs 4-38 and 4-39. Some
typical data of pulse velocity of HVFA concrete under water and air curing are
shown in Tables 4-27 and 4-28, respectively.
These results show that the ultrasonic pulse velocity curves can be divided
into three stages for all the HVFA concretes. In the first stage, the pulse velocity
increased rapidly for about one week, after which the reading showed an obvious
change under the two curing regimes. The second stage was a transition stage, and
the pulse velocity in this section exhibited a gradual increase or decrease about for 3
months. This changeable stage appears over the period between the end of the early
hydration and the pozzolanic activity period. There was seen during this period from
the preceding sections, a large variation of the swelling due to the formation of
ettringite ; growth of ettringite at early ages becomes more efficient in producing
swelling in the HVFA concretes and therefore, a little unstable stage of pulse velocity
development between the early hydration and the pozzolanic activity period is
understandable [77].
The third stage is a steady and stable phase with a gradual increase in pulse
velocity. The readings remained relatively constant with age at this stage as the
curing continued. From the data it can be inferred that the steady state implies that
the hydration products fill pore spaces and that the hydration reaction continues.
During the pozzolanic activity, the growth of hydration products continues to
progress gradually, and so, the pulse velocity reaches a stable progressive state.
109
Table 4-27 Ultrasonic pulse velocity on HVFA concrete ( Water curing)

( Unit : km/sec )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 3.57 3.47 3.62 3.51 3.65 3.65
7 4.25 4.16 4.22 4.23 4.24 4.28
28 4.23 4.47 4.53 4.31 4.5 4.53
91 4.55 4.61 4.66 4.55 4.65 4.66
189 4.68 4.71 4.71 4.69 4.72 4.69
364 4.78 4.73 4.75 4.77 4.78 4.75
504 4.78 4.77 4.74 4.79 4.77 4.75
5 ,---------------------------------------~
4.75
(,)
(])
~ 4.5
]
~
·0 4.25
o
.......
~
(])
-Ep. 4
3.75 450F 450SF 450S

350F 350SF 350S
Age (Days)
Figure 4-38 Variation of pulse velocity on HVFA concrete under water curing
Data from the results indicated that the ultrasonic pulse velocity was the
highest in water cured concretes. From Fig 4-38 it is seen that the presence of silica
fume in Mix-350SF and 450SF helped these concretes to develop a denser internal
structure from an early age. Mixes 350S and 450S with large amounts of fly ash also
110
developed a dense structure and their pulse velocity increased very rapidly from early
ages, also all they showed marginally better performance than the silica fume mixes,
but by about six months, all the mixes appeared to have more or less the same value
of ultrasonic pulse velocity. Beyond the age of six months, the pozzolanic reaction
activated by the fly ash and silica fume with the hydration of cement reduced the gap
between the different mixes. The long term pulse velocity developed for all the
HVFA concretes ranged from 4.74 Ian / sec to 4.79 Ian / sec at the age of504 days.
Under prolonged air curing all the HVF A concretes behaved somewhat
similarly, and showed a pattern of behaviour similar to that of dynamic modulus of
elasticity - a significant fall with age soon after the cessation of water curing, even
though the concretes had high strength. The results show that mixes 350S and 450S
performed best followed by those incorporating both fly ash and silica fume. At the
end of 504 days, the pulse velocity development of HVF A concretes ranged from
3.98 Ian / sec to 4.13 Ian / sec.
Table 4-28 Ultrasonic pulse velocity on HVFA concrete ( Air curing)

( Unit : km/sec )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 3.64 3.49 3.63 3.47 3.67 3.65
7 4.28 4.16 4.22 4.24 4.24 4.28
28 4.15 4.23 4.30 4.08 4.23 4.29
91 4.06 4.17 4.23 4.01 4.19 4.17
189 3.99 4.11 4.14 3.97 4.11 4.13
364 3.98 4.04 4.15 3.94 4.11 4.14
504 4.02 4.07 4.13 3.98 4.09 4.10
111
Chapter 4 Test Results and Discussion of Part I
112
5 r-----------------------------------____~
350F 350SF 350S 450F 450SF 450S
o o
•
4.75
u
(\)
~ 4.5
]
~.
'0 4.25
.....o
~
(\)
-3
~
4
3.75
56 112 168 224 280 336 392 448 504 560

Age (Days)
Figure 4-39 Variation of pulse velocity on HVFA concrete under air curing
The relative rates of development of pulse velocity of HVFA concretes cured

in water and air are illustrated in Table 4-29.
Table 4-29 Influence of air curing on pulse velocity

(Unit: %)
Age 350F 350SF 350S 450F 450SF 450S

( days)
14 97 96 97 96 96 97
28 98 95 95 95 94 95
91 89 90 91 88 90 90
189 85 87 88 85 87 88
364 83 86 87 83 86 87
504 84 85 87 83 86 86
Remarks : Water cured specimen is assumed 100%
112
The results indicate that exposure to ambient air of all HVF A concretes
caused a slight retrogression of ultrasonic pulse velocity with continued air drying.
The pulse velocity of air cured concrete is about 85 % of the average velocity of
water cured concrete at 504 days. The percentage gain of pulse velocity with
continued air drying is almost the same as the compressive strength development
under similar conditions. In general, the more dense and strong the HVF A concrete,
the lower is the value of the reduction. At the end of 504 days, the values of the pulse
velocity were in a manner similar to those of dynamic modulus of elasticity. The
average percentage loss on pulse velocity for all HVF A concretes is somewhat lower,
about 3 %, compared to the same concretes at the age of 28 days. The reduction in
pulse velocity is almost the same as the loss in dynamic modulus of elasticity, as was
also found by others [41].
A comparison of the influence of the total binder content on pulse velocity

under both types of curing conditions is shown in Figs 4-40 and 4-41. The results
show that the total binder content does not seem to have much effect on changes in
the pulse velocity under both types curing conditions. The reason may be explained
by the fact that, the pulse velocity is controlled primarily by the modulus of elasticity
of the material and not by it strength alone, because the composition of a concrete
does not affect its strength in the same way as it does its pulse velocity.
From the preceding data in Tables 4-23 and 4-24, similar rates of
development of dynamic modulus of elasticity due to total binder content were
observed. Therefore, for a given aggregate and a given richness of the mix, the pulse
velocity of the concrete is affected more by the modulus of elasticity, and
consequently, no unique relation between pulse velocity and compressive strength
exists [78].
113
5 5
-- - "_OM' -
~ 4.75 V ~ 4.75
] 4.5 ] 4.5
.~ 4.25 .~4.25
() ()
.Q 4 .Q 4
~ ~
~
~ 3.75
3.5 13~.~45OC I -~
~ 3.75
3.5
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
5 5
~ 4.75 ~ 4.75
]'" 4.5 ] 4.5

.~ 4.25 .~4.25
()
()
0 0
4 4
~ ~
~
~ 3.75
3.5
-~ ~ 3.75
3.5
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
Figure 4-40 Effect of total binder content on pulse velocity ( Water curing)
5 5
~ 4.75
V
~.. .-................. --- - ... ~ 4.75
'" '"
] 4.5 ] 4.5
.~4.25 .~ 4.25
() ()
.Q 4 .Q
~ ~
~
~ 3.75
3.5 13~~C450C I -~ ~ 3.75
3.5
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
5 5
~ 4.75 ~ 4.75
] 4.5 ] 4.5
~
.~ 4.25 ~ .~ 4.25 Ir' -
-~
() ~
~'
()
0 v v 0
4
4
~
'V
~
-~
~ 3.75
3.5 17 13:F45~F I ~
3.75
3.5 r 3!S4~SI
3.25 3.25
0 112 224 336 448 560 0 112 224 336 448 560
Days Days
Figure 4-41 Effect of total binder content on pulse velocity (Air curing)
11-'
Chapter 4 Test Results and Discussion o{Part 1 lI5
Table 4-30 represents some typical data of ultrasonic pulse velocity for the OPC
concrete mixes under both types curing conditions. The rate of gain of pulse velocity
as a % of 28 day velocity are is shown in Tables 4-31 and 4-32.
From the results, it can be seen that, under continued moist curing condition,
the percentage increase in pulse velocity from 28 days to one year is 3 % to 4 % for
all of normal OPC concretes. The corresponding average increase for the HVFA
concretes was 5 % to 13 %. The differences of between 350F and 350C, 450F and
450C mixes in pulse velocity at one year under water curing were about 10 % and 7
%, respectively.
At the end of one year, very little reduction in pulse velocity of OPC concrete
with air curing was observed. The differences between 350F and 350C, 450F and
450C mixes in pulse velocity at the end of on year under air curing were only 4 %
and 2 % respectively. Undoubtedly, the potential for pulse velocity development at
later ages for HVFA concretes exists, even if at a slow rate.
Table 4-30 Ultrasonic pulse velocity of normal ope concrete
Mix 350 kglm3 mixes 450 kglm3 mixes

km/sec Pulse velocity % of28 day Pulse velocity % of28 day
or velocity velocity
% (km/sec) J%) (km/sec) (%)
( days) curing curing curinA curing curing curing curing curing
1 4.60 4.58 97 97 4.55 4.55 97 97
7 4.66 4.66 99 99 4.66 4.65 99 99
28 4.72 4.70 100 100 4.70 4.71 100 100
91 4.80 4.73 102 101 4.78 4.73 102 100
189 4.82 4.76 102 101 4.83 4.73 103 100
364 4.86 4.71 103 100 4.88 4.65 104 99
115
Chapter 4 Test Results and Discussion o[Part 1 116
Table 4-31 Rate of gain velocity as a 010 of 28 day pulse velocity ( Water curing)
(Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 84 78 80 81 81 81
7 100 93 93 98 94 94
28 100 100 100 100 100 100
91 108 103 103 106 103 103
189 III 105 104 109 105 104
364 113 106 105 III 106 105
504 113 107 105 111 106 105
Table 4-32 Rate of gain velocity as a % of 28 day pulse velocity ( Air curing)
(Unit: % )
Age 350F 350SF 350S 450F 450SF 450S

( days)
1 88 83 84 85 87 85
7 103 98 98 104 100 100
28 100 100 100 100 100 100
91 99 99 98 98 99 97
189 96 97 96 97 97 96
364 96 96 97 97 97 96
504 97 96 96 98 97 96
4-7-3. Relationship between the pulse velocity and the dynamic

modulus elasticity
The relationship between the pulse velocity, V, and the dynamic modulus of
elasticity, ED, can be represented by the following equation:
116
Chapter 4 Test Results and Discussion o(Part I 117
v = a + b ED ------------------ ( 4-19 )
where V is in km / sec, ED is in GPa, and a and b are constants.
The binder content had shown similar influence on pulse velocity under both
curing conditions, the total binder content had a marked influence on the dynamic
modulus of elasticity. Further, the wet concretes had generally higher values of pulse
velocity and dynamic modulus of elasticity than those under dry curing. It is therefore
suggested that separate equation be used for each curing condition and total binder
content.
The variation of pulse velocity with dynamic modulus of elasticity for HVFA
concretes under the water curing is shown in Fig 4-42.
5.2 ~------------------------------~------------~
4.8
(.)
(l)
~
S
...!:oI::
4.6
..-
~ 4.4 •• •
.....
(.) o ••
o
'"a) 4.2
o
• ••
~ o
o
••
(l)
-S
~
4
o
3.8
3 3
350 kg/ m 450 kg/ m
3.6 o • Water curing
3.4 ~~~~~~~~~~~~~_L~~~~~~~~~~
30 35 40 45 50 55 60 65
Dynamic modulus of elasticity, GPa
Figure 4-42 Relationship between pulse velocity and dynamic modulus of

elasticity ( Water curing)
The relationship is based on measurements from 1 day to 504 days for moist
cured HVFA concretes. Regression analysis of the data shown in the figure produced
the following relationships:
117
Chapter 4 Test Results and Discussion of Part I 1J8
v = 0.86 + 0.45 ED 3
( 350 kg/m mixes) (r = 0.99 ) ------ ( 4-20 )
v = 0.84 + 0.44 ED 3
( 450 kg/m mixes) (r = 0.99) ------ ( 4-21 )
For the wet cured concrete, the two equations above are more or less
identical the same for the two total binder content concretes. The linear regression
shown in the figure has a correlation coefficient of 0.99. These data indicate a
significant correlation between ultrasonic pulse velocity and dynamic modulus of
elasticity for HVFA concretes at the 99 % confidence level.
In Fig 4-43, results of the one day to 504 days pulse velocity are plotted
against the dynamic modulus of elasticity for the air cured HVF A concretes.
Regression analysis of the test data shown in the figure produced the following
equations:
v = 0.44 + 0.63 ED ( 350 kg/m3 mixes) (r = 0.98 ) ------ ( 4-22 )
v = 0.39 + 0.62 ED ( 450 kg/m3 mixes) (r = 0.93 ) ------ ( 4-23 )
5.2 .----------------------~
Air curing 350 kg / m' 450 kg / rJ

5 o •
4.8
u
(\)
~ 4.6
]
~ 4.4
.....
u
o
Q) 4.2
~
(\)
~ 4
~
3.6
3.4 L-..l.-"/::::::..~--L....//___'-_L-__L_---L---L-J.-L---'-~__'_L___L__'___~__'_~_____'
32 34 36 38 40 42 44 46 48 50
Dynamic modulus of elasticity, GPa
Figure 4-43 Relationship between pulse velocity and dynamic modulus of

elasticity ( Air curing)
118
Chapter 4 Test Results and Discussion olPart 1 119
Again, the correlation factor between the two sets of equations is still high,
and the equation relating the two is approximately the same for the two total
cementitious content concretes. Irrespective of the curing condition, the pulse
velocity is controlled primarily by the dynamic modulus of elasticity of the concrete,
and ultrasonic pulse velocity increases with an increase in dynamic modulus of
elasticity.
It is worth pointing out, from the data for both types of curing conditions,
that concretes with the lower binder content have a marginally higher pulse velocity
than that with the higher binder content. This is in agreement with the concept, in
general, that the pulse velocity varies inversely to the density of concrete- V2 = ED / p,
where p is density.
4-8. CARBONATION OF HVFA CONCRETE
As mentioned in the literature review, one of the main causes of reinforcement

corrosion is the inadequate concrete cover, which is to protect the steel and keep it in
a passive state. This passivating film can be broken by the process of carbonation,
which reduces the level of pH adjacent to the steel to values that can no longer
maintain passivation. Once the steel has become depassivated, corrosion can occur if
moisture, chloride ions, and oxygen are present.
Because blended cements are widely used nowadays, it is important to know
the carbonation behaviour of concretes containing fly ash and silica fume. In HVF A
concrete, the replacement level is also a factor that affects carbonation rate. The
starting point in this section is the knowledge of the properties of the hardened
concrete resulting from the use of the various binder materials in so far as the
influence of these properties, physically or chemically, on carbonation is concerned.
The effect of the variables such as, mix proportions, and age of HVF A
concretes on carbonation were studied in a programme over 18 months under both
curing conditions. The relationship between carbonation depth and compressive
strength was also investigated.
119
4-8-1. Effect of curing condition and different replacement methods
The results of this investigation on carbonation for all the concrete mixtures under
both curing conditions are shown in Tables 4-33 and 4-34. For purposes of
comparison, the results of the normal OPC concrete on carbonation depth are also
included in the above table. Figs 4-44, 4-45, and 4-46 show the relationship between
the exposure period and carbonation depth of various HVFA and OPC concretes
stored in air. Figs 4-47 and 4-48 show typical examples of carbonation ofHVF A and
OPC concretes stored in air at various ages.
Table 4-33 Carbonation depth - 350 kglm3 mixes

(Unit: mm)
Age Mix 350C 350F 350SF 350S

( days)
Curing Water Air Water Air Water Air Water Air
curing curing curing curing curing curing curing curing
Max. 0.0 0.0 0.0 1.4 0.0 1.4 0.0 1.3
28 Min. 0.0 0.0 0.0 1.0 0.0 1.0 0.0 1.0
Ave. 0.0 0.0 0.0 1.3 0.0 1.3 0.0 1.1
Max. 0.0 1.8 0.0 6.4 0.0 5.7 0.0 2.8
90 Min. 0.0 0.0 0.0 3.2 0.0 2.0 0.0 2.0
Ave. 0.0 1.8 0.0 6.1 0.0 4.8 0.0 2.6
Max. 0.0 2.1 0.0 8.2 0.0 7.2 0.0 5.3
180 Min. 0.0 1.0 0.0 6.0 0.0 4.0 0.0 3.2
Ave. 0.0 2.0 0.0 6.5 0.0 5.1 0.0 3.5
Max. 0.0 2.3 0.0 10 0.0 8.6 0.0 6.5
360 Min. 0.0 1.5 0.0 9.0 0.0 6.0 0.0 5.0
Ave. 0.0 2.1 0.0 9.4 0.0 8.3 0.0 6.4
Max. --- --- 0.0 13.5 0.0 13.4 0.0 11.3
540 Min. --- --- 0.0 11.2 0.0 10.5 0.0 10.1
Ave. --- --- 0.0 13.3 0.0 13.2 0.0 11.1
120
16
350C 350F 350SF 350S
14 o. X 0 0
12
§
tCl)
10
Cl
d 8
.....0
~
d
0 6
~
U
4
2 o
0'
14 28 90 180 360 540
Age (Days)
Figure 4-44 Carbonation depth of concrete - 350 kglm3 mixes ( Air curing)
It is clear from the tables that, as expected, no trace on carbonation was

found in concretes stored continuously in water. The results for the air cured
concretes are, however, very different from those for the concretes stored in water.
Irrespective of initial water curing, the results show that the carbonation rate
increases as the environment becomes drier, and the values of carbonation depth
increase with increase in the exposure period. For both types of binder content
concretes, the normal ope concrete showed higher carbonation resistance, and
smallest carbonation depths compared to HVF A concrete. Further, with continued air
drying up to 18 months, concretes made of 450 kg/m3 showed a similar rate of
carbonation compared to the 350 kg/m3 mixes.
At one year, HVF A concretes that were water cured initially for 7 days
showed a significantly higher depth of carbonation than the corresponding normal
ope concretes for both types of binder contents. The depth of carbonation in normal
ope concrete is, on average, only 25 % of that of the HVFA concrete. The higher
carbonation of HVF A concrete is attributed to greater permeability of the surface
concrete containing fly ash.
121
Table 4-34 Carbonation depth - 450 kglm3 mixes

(Unit: mm)
Age Mix 450C 450F 450SF 450S

( days)
Curing Water Air Water Air Water Air Water Air
curing curing curing curing curing curing curing curing
Max. 0.0 0.0 0.0 1.4 0.0 1.5 0.0 1.3
28 Min. 0.0 0.0 0.0 1.0 0.0 1.0 0.0 1.0
Ave. 0.0 0.0 0.0 1.3 0.0 1.2 0.0 1.1
Max. 0.0 1.3 0.0 5.7 0.0 3.9 0.0 2.1
90 Min. 0.0 1.0 0.0 3.3 0.0 2.0 0.0 1.0
Ave. 0.0 1.3 0.0 5.3 0.0 3.7 0.0 1.5
Max. 0.0 1.9 0.0 7.3 0.0 6.1 0.0 5.2
180 Min. 0.0 1.7 0.0 4.0 0.0 3.5 0.0 2.8
Ave. 0.0 1.8 0.0 5.5 0.0 4.5 0.0 2.7
Max. 0.0 2.1 0.0 9.3 0.0 7.4 0.0 6.3
360 Min. 0.0 2.0 0.0 4.0 0.0 3.5 0.0 5.0
Ave. 0.0 2.0 0.0 9.2 0.0 7.2 0.0 5.1
Max. --- --- 0.0 13.5 0.0 12.5 0.0 11.0
540 Min. --- --- 0.0 11.0 0.0 11.1 0.0 11.0
Ave. --- --- 0.0 13.3 0.0 12.2 0.0 11.0
122
16
450C 450F 450SF 450S
14 D I •
12
~
t Q)
10
0
d 8
.....0
~
d
0 6
~
u
4
0 1
14 28 90 180 360 540

Age (Days)
Figure 4-45 Carbonation depth of concrete-450 kglm3 mixes ( Air curing)
16 16
~ 14 I 350C 450C I ~ 14
.g, 12 .g' 12
8'10 8'10
§ 8 § 8
.~
6 ".g 6
=
0
4
=
0
..D 4
~ t;J
u 2 u 2
0 ~ 0
14 28 90 180 360 14 28 90 180 540
Days Days
16 16
~ 14 350SF 450SF
-<7 ---
§14
.g' 12 ..d' 12
8'10 Q..1O
8
§ 8
"+=I
=
0
8
.~
~
t::
6 6
t::
0 0
... 4
..0
~
..0
t;J
4
u 2 u 2
0 0
14 28 90 180 540 14 28 90 180 540
Days Days
Figure 4-46 Effect of total binder content on carbonation depth ( Air curing)
123
Chapter 4 Test Results and Discussion of Part 1 12-1
,
'"
6-
I
... • •l
"• -
.t/ "
,
450f -I RO
Figure 4-47 Carbonation of specimens stored in the laboratory-180 days
50I ,- IX i\!
Figure 4-48 Carbonation of specimens stored in the laboratory-540 days
12-l
Further, under air curing, it is also clear from the above data that concrete,
Mix-350F and Mix-450F, exhibited higher carbonation depths than those obtained in
the concretes incorporating both fly ash and silica fume.
At the end of 18 months, for all HVFA concretes under air drying, the depth
of carbonation was a great deal higher than the values received at one year. For
specimens stored in air, differences in carbonation depth were found for concrete
mixes 350F and 450F, and the mixes containing fly ash as a partial replacement for
both sand and cement.
The influence of fly ash and silica fume on carbonation of concretes found in
this study is in agreement with the results of other investigators [27,44,79]. The
increase in the carbonation depth of HVFA concrete compared to ope concrete may
be due to the individual or combined effect of their greater surface porosity, and its
smaller concentration of carbonatable constituents [5]. This indicates that the
addition of such mineral admixtures to concrete which does not form proper matrix
films upon coalescence, may rather increase the risk of carbonation. Kokubu &
Nagataki also observed that the carbonation was distinctly increased due to raising
the fly ash replacement ratio, among other things [80].
The important influence of curing condition on carbonation is clearly seen in
those results. As for depth of carbonation, the depth increases the longer the period
of exposure in air, which is considerably higher than that of the concrete exposed to
moist curing. The results are also relevant for the strength development as well as
dynamic modulus of elasticity, ultrasonic pulse velocity for high strength concretes. If
the concrete matrix is porous due to carbonation, strength reduction generally
occurs. As discussed in preceding sections, strength development of HVF A concrete
in case of curing in water increases with age. When cured in air, although there is an
increase in strength seen from the age of four weeks to one year, a trend of decline is
recognised. Similar trends were observed for the dynamic modulus of elasticity and
pulse velocity. These results indicate that high strength alone is not sufficient to
prevent carbonation, moisture is also needed to slow down the carbonation process.
These findings are in accordance with the results of other investigators [48,81,82,83].
Many of them suggest that an increase in curing time of the concretes with admixture
125
Chapter 4 Test Results and Discussion olPart 1 126
reduces the difference with ope concrete regarding the rate of carbonation, as well
as other physical and mechanical properties of concrete.
4-8-2 Carbonation rate coefficient
In literature [83,84] the rate of carbonation is found to be calculated as a function of

exposure time, water / binder ratio, strength development, etc.
It is thus possible to express the depth of carbonation as :
e =a +K Ji ---------------- ( 4-24 )
where C is the carbonation depth in mm, K is the carbonation rate coefficient, t IS
the time of exposure in sec, and a is constant.
Regression analysis to define the relationship between exposure period (Ji )

and the depth of carbonation is shown in Figs 4-49 and 4-50. The parameters, a, and,
K, and the coefficients of determination, ( r), are given in Table 4-35.
Table 4-35 Estimated parameters of equation C = a + K..Ji (t = sec)
Mix a K r
350e -0.16806 0.00047 0.81882
350F -0.61110 0.00185 0.94733
350SF -0.72942 0.00161 0.96740
350S -1.06045 0.00128 0.98733
450e -0.38068 0.00047 0.90919
450F -0.96996 0.00181 0.96348
450SF -0.79531 0.00142 0.98826
450S -0.92199 0.00101 0.96652
126
Chapter 4 Test Results and Discussion o{Part 1 127
16
350C 350F 350SF 350S
14 0 )( 0 0
12
§
~ 10
x
fr
0
§ 8
.+=!
Cd
§ 6 o
x
~
u
4 o
2 o o o
0
1,000 2,000 3,000 4,000 5,000 6,000
Time, Sec
Figure 4-49 Exposure period versus carbonation depth of concrete-350 kglm 3

mixes
16
450C 450F 450SF 450S
14 o •
12
~
o£3" 10
fr
0
§ 8
.+=!
Cd
§ 6
~ + + •
u
4
2 o
o ~~~ __- L_ _~_ _~_ _ _ _~_ _L -_ _~_ _-L--~--~
1,000 2,000 3,000 4,000 5,000 6,000

Time, Sec
3
Figure 4-50 Exposure period versus carbonation depth of concrete-450 kglm
mixes
127
These parameters are valid for exposure times up to one year. All the HVF A
concretes show higher carbonation rate coefficient, K, in comparison to normal ope
concretes because of the reasons mentioned earlier. Further, concrete with the 350
3
kg/m binder content had a higher carbonation rate coefficient than that of the 450
3
kg/m mixes. Again, the concretes made of 350F and 450F exhibited the poorest
results.
4-8-3. Relationship between the carbonation depth and compressive

strength
In general, the behaviour of concrete is always based on the assumption that the
ambient medium, i. e., air does not react with the hydrated cement. In fact, structural
concrete is always exposed to the atmosphere which contains high levels of carbon
dioxide. In the presence of moisture, the carbon dioxide reacts with the hydrated
cement, and this results in carbonation taking place through the pore system in the
hardened cement concrete, which may result in a decrease in strength. The rate of
carbonation depends on the atmospheric carbon dioxide, cement content, relative
humidity, temperature, and the porosity of the concrete, the latter depending upon
the type of cement, the water / binder ratio, and the degree of hydration.
The relationship between the depth of carbonation and compressive strength

for dry cured HVFA concretes, at ages of 180 days, 360 days, and 540 days is shown
in Fig 4-51.
Regression analysis of the data shown in the figure produced the following
relationships:
fc = 87.39 - 6.80 e ( 180 days) (r= - 0.95 ) ------ ( 4-25 )
fc = 87.40 - 4.37 e ( 360 days) (r= - 0.98 ) ------ ( 4-26 )
fc= 102.9 - 3.99 e ( 540 days) ( r = - 0.85 ) ------ ( 4-27 )
where fc is the compressive strength in MPa, and e is the carbonation depth in mm.
128
It is often said that the rate of carbonation is simply a function of the strength
of concrete. As shown in the figure, there is a reasonable inverse correlation between
the depth of carbonation, C, and compressive strength, fe, of HVF A concrete : in
other words, the depth of carbonation of concrete increases with decrease in strength.
As for depth of carbonation, the depth increases the longer period the of time
left in air. At the end of 540 days, although the carbonation and strength relationship
shows a lower correlation coefficient, the data indicate still significant correlation at
the 99 % confidence level between the depth of carbonation and compressive
strength. On the basis of the present studies, of the factors affecting carbonation,
strength appears to be an important factor.
90 ~----~--------------------------------------,
80
0
c;I
p. 180day
:E~ 70 Air
t
~
rJJ 60
+ + 0
0
360day
Air
.-
Q)
>
t il
til
Q)
+
+ 0
0
0
0
540day
+
E' 50
'""' Air
0
U
0
0
8
40
Carbonation Depth, mm
Figure 4-51 Relationship between carbonation depth and compressive strength
129
4-8-4. Comparison with published data
Comparison of carbonation depth with those of other investigators [27,44,48,59] is

shown in Table 4-36. Data in table clearly shows that there is a wide difference in
carbonation depths. Due to the different test parameters, it is difficult to analyse the
test results from the data and give a definite conclusions. However, it is obvious from
the results that fly ash concretes having higher cementitious content, low water /
binder ratio and moisture, are most likely to well resist from carbonation. Meanwhile,
the importance of early strength is also clearly indicated by the data.
Table 4-36 Comparison of carbonation depths with published data
Researchers Binder Fly ash ~/E 28 day Carbonation Test age

content content strength depth
3
II kg/m ) (%) (MPa) (mm) (Months)
Malhotra & Bilodeau 370 58 0.32 35.0 4.0 31
[44] " " " " 7.0 50
" " " " 8.0 90
Swamy & Mahmud 266 50 0.62 23.7 23.6 ~ 27.2 33
[59] 348 " 0.42 46.6 13.2 ~ 17.9 "

466 " 0.32 63.3 6.3 ~ 11.1 "
Langleyet al 400 55 0.27 44.0 3.0 24
[27] 225 56 0.49 22.1 10.0 "

Kawamura & Haque 250 30 0.55 35.0 7.5 48
[48] 350 " 0.49 45.0 5.0 "

Author 350F 57 0.40 28.4 13.3 18
350SF 51 " 29.9 13.2 "

350S 97 " 38.9 11.1 "
450F 56 " 28.6 13.3 "
450SF 51 " 38.1 12.2 "
450S 97 " 39.3 11.0 "
130
4-9. CONCLUSIONS
This chapter has examined the effect of curing condition on the engmeenng
properties of HVFA concrete with and without silica fume. The following
conclusions are based on the results and analyses presented in this chapter.
1. For the HVFA concrete investigated, continuous moist curing gave higher
strength, dynamic modulus of elasticity, and ultrasonic pulse velocity than
prolonged exposure to air drying.
2. Under moist curing, there was a continuous increase in compressive strength

with age. The strength at 18 months expressed as a percentage of the 28 day
strength varied between 157 % and 201 %.
3. Air curing preceded by 7 days moist curing is beneficial to the flexural strength
development of HVFA concretes. At 90 days, concrete under this condition
exhibited the highest flexural strength.
4. A significant correlation between flexural strength and compreSSIve strength

exists for HVFA concrete at the 99 % confidence level.
3
5. In general, increasing the binder content of the concrete from 350 kg/m to 450
kg/m3 of binder content had little effect on strength, but did yield a large benefit
in terms of dynamic modulus of elasticity, and pulse velocity in most of the
HVF A concretes.
6. Under the same conditions, concretes made with both fly ash and silica fume had
engineering properties in most respects which were as good as, or superior to,
those made by fly ash alone as cement replacement. The use of fly ash to replace
both cement and sand was able to combine the benefits of both separate
replacements.
7. Air curing causes a substantial reduction in the engineering properties of HVF A

concretes. For example, the compressive strength of dry cured HVF A concrete is
about 83 % of the strength of wet cured concrete. The average percentage loss
131
of dynamic modulus of elasticity, or pulse velocity for all HVF A concretes was,
on the other hand, very low, only about 3 %, when compared to the same
concretes air cured for 28 days.
8. The deterioration of concrete also influences the cement paste in the interior of
HVF A concrete, and has a close relationship with carbonation. The phenomenon
of a decline in engineering properties accompanying increase in depth of
carbonation is also seen in normal portland cement concrete.
9. Although the early strength of concrete made with HVFA was found to be low,
those at ages beyond 28 days showed a continued rate of development of
hydration, superior to that of the normal concrete. The HVFA concrete in this
study has good potential for use in structural and mass concrete applications.
10. The swelling in the HVFA concretes was associated with formation of ettringite.
The ettringite formed in the early ages of hydration was more efficient in
producing swelling. This suggests that swelling began at a lower critical degree
of hydration ; the slower rate hydration of the HVFA concrete probably resulted
in slightly more swelling due to adsorption of water.
11. Drying shrinkage did not correlate with total porosity but correlated reasonably
well with the volume of micropores plus mesopores.
12. The drying shrinkage strain measured on all HVFA concretes was comparable to
Ross's formula, and was about 95 % of the average percentage of the latter.
3
13. The values of static modulus of elasticity, Be, of 450 kg/m mixes under water
curing are higher than the calculated values according to ACI building Code
318-89.
14. The dynamic modulus of elasticity varies with changes in the total binder content
similarly to normal portland concrete mixes, and is curvilinear in its relation to
compressive strength.
132
15. There is no physical relation between the ultrasonic pulse velocity and strength.
The pulse velocity was found to be very sensitive and related to the density of
concrete, and linearly related to the dynamic modulus of elasticity. The total
binder content does not seem to have much effect on pulse velocity under both
types curing conditions. Under moist curing, all of HVF A mixtures appeared to
have more or less the same value of pulse velocity beyond about six months.
16. Overall, the depth of carbonation was found to be related to the compressive
strength of concrete, with and without fly ash. Generally, the HVF A concrete
carbonated to somewhat higher depth compared to the normal portland concrete
exposed to air curing.
17. High strength alone is not sufficient to prevent carbonation, moisture is also
needed to slow down the carbonation.
133
Chapter 5 Microstructure of Hydrated Cement Pastes 13.+
CHAPTER FIVE
MICROSTRUCTURE OF HYDRATED CEMENT

PASTES
5-1. INTRODUCTION
The engmeenng properties of normal ope concrete, and, of HVF A concrete,

containing silica fume were presented in Chapter 4, together with a discussion of the
influence of fly ash and silica fume on the properties of HVF A concrete. In practice,
however, most of the important properties of hardened concrete are related to the
quantity, and the characteristics, of the various types of pores and hydration products
in the concrete. Indeed, the engineering properties, such as strength development,
drying shrinkage, swelling, and permeability are ,directly influenced and controlled by
I
the relative amounts of the pores and the hydration products.
There is, thus, a great impetus to study the microstructure of such a concrete
in order to examine the changes of hydration of HYF A concrete progressing with
time. To a lesser extent, the influence of the characteristics of microstructure are also
discussed but, it is useful to review the pore structure and mineralogical compositions
of concrete.
5-2. CLASSIFICATION OF PORE SIZE IN HYDRATED

CEMENT PASTES
The pores formed in fresh concrete are either water-filled or gas-filled. At the initial
mixing period, the cement grains, and the conglomerates in the fresh pastes, are
separated individually by the mixing water, which forms the water-filled space. The
134
Chapter 5 Microstructure of Hydrated Cement Pastes 135
water-filled space in the freshly mixed cement paste represents space that is available
for the formation of cement hydration products.
As hydration proceeds, the volume of this space, which has a bulk volume
larger than that of the original unhydrated cement, is continually reduced by the
formation of the hydrate products that form a continuous matrix and bind the residual
cement grains together over a period of time. At any time, that part of the original
water-filled space not occupied by hydration products constitutes part of the pore
system of the paste, and generally speaking, that part with the largest pores sizes.
Hydration reduces the size and volume of this capillary space. If the original capillary
space is small, the bulk volume of the gel eventually will be sufficient to fill this space
and produce a paste free from capillary space.
In general, the volume of the water filled spaces in fresh paste must exceed
the absolute volume of the cement, or some of the original cement must remain
unhydrated. For this reason, the hydration products are not sufficient to fill the
original water space in the paste, even after complete cement hydration, and pastes
having a relatively large volume of capillaries will be produced.
Studies on the hardened cement pastes have shown that the size of its largest
pores depends principally on the degree of hydration, varying from about 111m for
young pastes in which the cement hydration has progressed only a little, to about
O.ll1m in mature pastes that are more completely hydrated.
As the spaces become filled with hydration products on continued curing, a

time may come, if the original water / cementitious ratio was small enough, that most
of the capillary spaces become isolated cavities, separated from each other by the gel
that is the hydration product. This gel is itself a porous solid, most of which is the
calcium silicate hydrate ( C-H-S gel ), and the non-solid part.
As mentioned in section 2-4, the C-S-H gel is a rigid substance, made up of

colloid-size particles, and has a characteristic porosity of 28 %. The average width of
the pores involved is about 15 A and they are known as "gel pores" [13]. The gel
pore sizes are too small to allow the nucleation of new hydration products. If a paste
135
has such a low water / cementitious ratio that not all the cement can even become
hydrated, the gel will fill completely all the space not occupied by unhydrated cement,
and will contain porosity.
Measurement and classification of pore size is surrounded by experimental

difficulties and by controversy. There are two main problems. First, the paste must be
dried in order to make measurements, and drying almost certainly changes the pore
structure. Second, a definite geometrical shape of pore must be assumed while
S.E.M. micro-graphs show that the pores from a network of very irregularly shaped
pores.
Pore sizes have been classified in different ways by Powers, Feldman and
Sereda, Diamond et al , IUAPC, and Bentur et al [85,86,87,88]. A comparison of
pore classification is shown in Fig 5-1. Table 5-1 shows a classification of porosity in
cement paste by Powers model, it can be seen that there is an enormous range of pore
sizes, from 10 pm to less than 0.005 pm ( 0.5 nm) in diameter.
,----
Reference
I Gel pores I Capillary pores Powers

,
i
Diamond
I Interlayer
I
I Crystalline I Intergel I Macropores Feldman & Sereda
I Micropores I Mesopores I Macropores IUPAC
10 100 1000 10000

0
Pore Diameter, A
'-.
Figure 5-1 Comparison of pore classifications [85,86,87,88]
136
Table 5-1 Classification of pore sizes in hydrated cement pastes [13]
Designation Diameter Description Role of water Paste properties

affected
Capillary pores 10 - 0.05 pm Large capillarie~ Behaves as Strength
bulk water permeability
50 - 10 nm Medium Moderate Strength
capillaries
surface permeability
tension shrinkage
forces at high
generated humidities
Gel pores 10-2.5nm Small (gel) Strong surface Shrinkage to
capillaries tension 50%RH
forces
generated
2.5 - 0.5 nm Micropores Strongly Shrinkage
adsorbed water creep
<-0.5 nm Micropores Structural Shrinkage
water creep
involved in
bonding
5-3. INFLUENCE OF POROSITY ON STRENGTH
All porous materials have this same inverse dependency between strength and
porosity. Both theory and experiment lead to relationships that are either parabolic or
exponential between the two. The effect of porosity on the strength of hydrated
cement paste has been studied widely. Many investigators [89] have recognised that
porosity is an important parameter in determining strength of hardened cement
pastes, and its effect is expressed by a power function of the type :
fc = fo xn --------------- ( 5-1 )
l37
where fc is strength of the paste, ~ is a constant representing the intrinsic strength of

the cement gel ( i. e., the strength at a gel/space ratio of 1.0 ), and n is a constant,
which has values in the range 2.6 to 3.0, depending on the characteristics the cement.
X is the gel/space ratio, which is defined as :
0.68a
X = --------------- ( 5-2 )
0.32a+%
where a is the degree of hydration, and w / c is the water / binder ratio.
This knowledge has been used in preparing cement paste; cement is mixed
with only enough water to make it workable, thereby minimising the water remaining
in pores after the paste hardens.
Relationship between the strength and gel/space ratio is shown in Fig 5-2.
The relationship is based on results tested from 2 in. mortar cubes [69]. A linear
relation between strength and porosity, within the range of the latter between 5 and
28 %, was established by Rossler and adler [90].
Watson [91] has pointed out the importance of porosity and derived a
formula which expresses the relationship between strength and V s / Vp , the solid /
pore volume ratio, where V s and Vp are the volume fractions of the solid material and
pores respectively as determined by helium pycnometry. Meanwhile, Feldman and
Beaudoin [92] have emphasized that the porosity is a major factor in controlling
strength, and modulus of elasticity of normal OPC concrete systems. The relation
between the strength of mortar based on volume of pores larger than 20 nm in
diameter is shown in Fig 5-3 [93].
Consequently, in addition to total porosity, the effect of pore size distribution

on strength must be considered. Generally, at a given porosity, small pores lead to
higher strength of the cement paste. Goto and Roy [94] also report results for cement
pastes which agree well with experimental data from Bentur [71].
138
Chapter 5 lvficrostructure o{Hydrated Cement Pastes 139
140
&! 120 •
~
rA'
.B::s 100
U
-.
.-
d 80
--~
C"'I
60
.......
V)
~
0
40
t -t:.
83
-=
rF.l 20 •
o L -_ _~_ _ _ _~_ _ _ _~_ _ _ _~_ _ _ _~_ _~L- __ ~ ____ ~
0.2 0.4 0.6 0.8 1

Gel/Space Ratio
Figure 5-2 Relation between the compressive strength of mortar and gel/space
ratio [69]
60.---------------------------------------------
50
• • •
•
• •
10 L-~ __~__~___ L_ _~_ _~_ _~~_ _~_ _~_ _~~
9 10 11 12 13 14 15
Porosity-per cent
Figure 5-3 Relation between compressive strength of mortar and calculated

from the volume of pores larger than 20 nm in diameter [93]
139
Chapter 5 JvIicrostructure of Hydrated Cement Pastes 1-10
5-4. INFLUENCE OF POROSITY ON DRYING SHRINKAGE
The pore size distribution of a paste is considered important in determining drying

shrinkage behaviour. For example, the addition of calcium chloride, increases
irreversible shrinkage on first drying, and a higher contribution to shrinkage
attributable to the capillary component [70,71].
Young and co-workers [95] have attempted to correlate microstructure with

shrinkage. Various microstrutureal parameters were measured and the shrinkage
behaviour studied ; however, it was found that no one parameter could correlate with
shrinkage and porosity. Another approach using composite structure factor showed
good correlation, and it was also found that this relationship can be applied to
portland cement pastes, with reasonable success. The polysilicate content was also
used as a microstructual parameter in a study linking shrinkage with microstructual
properties [70,71].
From above analysis and further studies, Young and co-workers [96]
conclude that shrinkage is the result of two components which are superimposed
upon one another :
1. Capillary drying shrinkage - loss of moisture from the "pore component", meso-
pores: 25 A ~ 300 A, which is assumed to be due to capillary effects; and
2. Gel drying shrinkage - loss of moisture from the "C-S-H component" which
includes the C-S-H particles and the pores smaller than 25 A (micropores ).
5-5. INFLUENCE OF POROSITY ON PERMEABILITY
Establishing the pathways and mechanisms for fluid flow through cementitious
materials is an important part of the development of the understanding of mechanisms
of ionic transport. The permeability is affected by many factors such as hydrostatic
pressure, osmotic effects, pore differentials, and temperature differentials. The
140
Chapter 5 Microstructure o(Hydrated Cement Pastes ].11
permeability of concrete can be measured by determining the rate of flow of water or

air through a concrete.
Flow in concrete paste and concrete obeys Darcy's law :
dv h
- = K - --------------- (5-3 )
dt x
where v is the velocity of flow, K is the coefficient of permeability, h is the

pressure head, and x is the thickness of the specimen.
In a fresh paste, the flow of water is controlled by the size, shape, and
concentration of the original cement particles. With the progress of hydration, the
permeability decreases rapidly because the gross volume of gel ( including the gel
pores) is approximately 2.1 times the volume of the unhydrated cement, so that the
gel gradually fills some of the original water-filled space. As hydration proceeds, the
capillary network is cut offby C-S-H gel, and then the capillary pore system becomes
discontinuous. Flow is now controlled by micropores and this is very slow. This is
accompanied by a continuous decrease in coefficient of permeability, "K".
In general, water can flow more easily through the capillary pores than
through the much smaller gel pores. The cement paste as a whole is 20 to 100 times
more permeable than the gel itself [69]. In a mature paste, the permeability depends
on size, shape, and concentration of the gel particles, and on whether or not the
capillaries have become discontinuous. It is worth to note that low-pressure steam
curing shows a slight increase in permeability [97].
If the specimen is not cured well, drying shrinkage may rupture some of the
gel between the capillaries, and thus result in larger permeability coefficient values.
On the contrary, the specimen cured sufficiently in water has more moisture, and it
has denser structure because of more hydration than in specimen with drying.
Goto and Roy [94] presented some experimental results showing the
existence of a relationship between permeability, and pore size distribution. The
relationship between porosity and permeability was also demonstrated by Powers
141
who confirmed that the coefficient of permeability decreases with a decrease in the
percentage of capillary porosity.
Mehta [12] showed that additions ofpozzolanic and cementitious admixtures,

such as fly ash, to normal ope cement were instrumental in causing pore refinement
or transformation of large pores into fine pores - a process which had a far reaching
influence on the permeability of the hardened cement pastes.
5-6. INFLUENCE OF POROSITY ON ABSORPTION
The term "absorption" is usually applied to concrete in regard to the weight gain of
partially dried specimens upon contact with or immersion in water. However, water
can enter the pores, the amount of the absorbed water depending primarily on the
abundance and continuity of the pores in the particle, whereas the rate of absorption
depends on the size and continuity of these pores.
During absorption, it may be considered that the larger capillary spaces in the
paste are the first to be wetted, with the finer gel pores perhaps next, their rate of
absorption being slow because of their low water permeability. Thus, these features
of the pores are more important than merely their total volumes as reflected by the
magnitude of water absorption.
Although there is no clear-out relation between the strength of concrete and

the water absorption of aggregate used, the pores at the surface of the particle affect
the bond between the aggregate and the cement paste, and may thus exert some
influence on the strength of concrete. Because the materials of aggregates always
contain more or less pores inside, practically all dry particles are capable of absorbing
water. Meanwhile, the coarse pores in aggregate can become nearly filled with water,
when a relatively high degree of saturation is established in the paste surrounding the
aggregate.
142
Chapter 5 Microstructure ofHydrated Cement Pastes 143
It may be noted that gravel has generally a higher water absorption than
crushed rock of the same petrological character, because weathering results in the
outer layer of the gravel particles being more porous and absorbent [69].
From further studies, it is generally held that the water absorption of mature
concrete is not a simple function of its porosity, but depends also on the pore size,
distribution, water / cementitious ratio, aggregate characteristics, air content, cement
type and fineness ( particularly at early ages ), specimen size and shape, method of
surface preparation ( cast, broken ), degree of hydration, continuity of the pores, and
so on.
5-7. MICROSTRUCTURE OF HARDENED CEMENT PASTE
Research into the hydration of cement has been directed, on the one hand, towards
the phenomenon of engineering characteristics in term of microstructure and
porosity, and, on the other, to the way in which the chemical composition of the
cement and its hydration products can affect its properties or be modified to improve
them.
A considerable amount of knowledge of the mechanism of hardened cement

paste has been built up as a result of studies on the microstructure. In a recent
review, Skalny, Jawed and Taylor have considered various mechanisms proposed for
the hydration of cement, and compared them with new experimental approaches. The
kinetics of reactions have been summarised by Kantro, Weise & Brunauer and by
Kondo & Veda. The structure of the calcium silicate hydrates has been reviewed by
Taylor and the structure of calcium aluminate hydrates by Turriziani and Schweite &
Ludwig. Powers and Brownyard, who studied the microstructure of hardened cement
in relation to its physical properties, hypothesised that gel particles 100 A in diameter
are separated by a water film 6 A thick. Subsequent investigations, however,
indicated a structure with a unit layer ~ 10 A thick [98,99,100].
143
Chapter 5 Microstructure ofHydrated Cement Pastes ]./-1
According to Powers-Brunauer model, the cement gel itself contains gel

pores ~ 15 A trapped between folded sheets of crystalline layers of Ca, Si, and Al
atoms. These layered sheets some 20 ~ 30 A thick may touch in places and be as
much as 500 A x 1 /lm in extent. Water held between the layers is considered tightly
bound water, part of the crystal lattice, but the water in the gel pores can be more
readily lost, and it is this water that accounts for shrinkage and creep.
Chapter two briefly dealt with the hydration reaction of portland cement in
general, and cements differing in chemical composition may exhibit different
hydration products when hydrated.
Upon addition of water the cement begins a series of hydration reactions, the
most important of which is the formation of a poorly crystallised hydration product
called calcium silicate hydrate gel ( C-S-H gel ), and calcium hydroxide ( CH ) by the
calcium silicates. Other hydration products involving iron, aluminium and sulphur
constituents are also formed. Thus, this section deals with the hydration products of
portland cement, and its crystal structure, and properties relevant to the present aim
of the investigation.
5-7-1. Calcium silicate hydrate gel (C-S-H gel)
The C-S-H gel formed in cement paste has been thought to have a highly disordered
structure with a small particle size ( dimension small than 150 A ), and a high surface
area [101]. It is termed "gel" because its dimensions are in the colloid range ( less
than 1000 A ) and is poorly crystalline, being essentially amorphous. Jenning and
Pratt [85,102] have shown the thin sheets or foil-shaped C-S-H structures in the
saturated state from T .E.M. ( Transmission Electron Microscope) observations.
Based on the C / S and H / S ratio so far reported, the composition of the C-

S-H gel may be represented by the ternary diagram of Fig 5-4 [103]. From the figure,
there are three main varieties of well crystallised tobermorites. All three varieties are
lamella, and have the basal spacing 9.3 A, 11.3 A, and 14 A. There are also semi-
crystalline tobermorites known as C-S-H ( I ) and C-S-H ( II ) where the C / S ratio

varies over the range 0.5 ~ 3.0. C-S-H ( I ) is usually observed as crumpled foils and
C-S-H (II) appears mostly as fiber.
The similarity between C-S-H gels and tobermorite has been demonstrated by
thermal analysis, electron and X-ray diffraction, and infrared spectroscopy. Results of
X-ray diffraction analysis, however, remain the main basis for this association
between C-S-H gels and tobermorite. The C-S-H gels give three broad lines
corresponding to 3.05 A, 2.8 A, and 1.82 A reflections of crystalline tobermorites
[103].
H
I **** IGeI in Paste
CSHI
9T
SL-----------~~----~~~------~c
C/ s= 1.3 C/ s= 2.3
L_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~
Figure 5-4 The composition of the calcium silicate hydrates [103]
The chemical compositions of C-S-H gel are quite diverse, especially when
other chemical compounds are added, and it is believed that solid state substitutions
will occur, which has been confirmed by electron beam micro-analysis [104].
1~5
Recently, the C-S-H chemical composition studies have concentrated on

silicate structure analysis. C-S-H can be considered to be a polydisperse material with
a relatively low degree of silicate polymerisation [86,105,106,107].
The physical characteristics of C-S-H are also diverse, and these characters
dependent on the technique of measurement, the sample preparation procedure and
the environmental condition. The mechanical properties of the hydrated gel are
described by Powers and Brunauer, who presumed that the particles are to be held
together by Van der Waals forces. Swelling on exposure to water is explained by the
individual particles separating due to layers of water molecules existing between
them, and this model also includes the existence of some chemical bonds between the
particles to explain the limited swelling nature of the materials. Feldman and Sereda
have postulated that the interlayer regions of C-S-H play an important role in
shrinkage, particularly irreversible shrinkage.
The exact nature of the C-S-H microstructure is still uncertain, due mainly to
the lack of direct structure determination techniques [108]. However, the C-S-H gel
is still considered to be primarily responsible for the strength and cementing
characteristics of the concrete. Also, it is now believed that the removal of water by
any drying process will more or less change the C-S-H structure, and that the dried
paste will be different from the saturated paste. The various morphologies and the
various compositions of C-S-H also cause problems in understanding.
5-7-2. Calcium hydroxide (CH)
A major constituent of hydrated portland cement is generated by reaction of alite and

belite with water. It is a well-crystallised solid containing hydroxyl ions, but no water
molecules.
Calcium hydroxide dehydrates to calcium oxide at about 400°C, in D. T. 1\

and this gives rise to a strong endothermic peak at about 550°C. Calcium hydroxide
in cement pastes may vary in morphology, being found as small equidimensional
146
crystals, thin plates, elongated crystals, flakes, or variation of these forms [87]. The
phase of calcium hydroxide gives six strongest X-ray diffraction peaks corresponding
to the 2.63 A., 4.90 A., 1.93 A., 1.80 A., 3.11 A., and 1.69 A. Its morphology is
particularly affected by admixtures and by the temperature of hydration.
5-7-3. Calcium sulfoalumimate hydrates
The calcium sulfoaluminates are a relatively minor constituent of a cement paste,

making up perhaps 10 to 20 % by volume. Thus, they are generally ignored due to
their correspondingly minor role in the microstructure of the hydrated cement past~s.
This hypothesis may not be entirely justified, however, since these compounds do
contain relatively large amount of quite unstable ettringite, and the hexagonal
hydrates.
The ettringite takes the form of needles, and because of its relatively low
specific gravity ( 1.70 ) occupies a correspondingly large volume. Excessive
formation of ettringite may lead to tensile stresses in the concrete, which may lead to
cracking of the concrete.
The morphology of hexagonal hydrates, C4AH 13 , C2AHg, can be clusters of

small, plate crystals instead of well hexagonal crystal. The X-ray basal spacing of
C4AH13 is 7.9 A . There is an unsettled dispute as to whether the composition of this
phase is really C4AH13 or C4AH 12 . The compound may be formed in hydrated
portland cement high in alumina and low in sulphate. The water molecules are loosely
held in the layer structure of the hexagonal hydrates. Drying can remove some of the
water between the layers forming new hydrates. Thus, loss of water might possibly
cause a bulk contraction of the crystal.
The hexagonal calcium aluminate hydrates in solution convert slowly to a

more stable cubic hydrate C3Aai. C3 AfL, is the chemically stable final product at
room temperature, and is a member of the hydro-garnet solid solution series. It
147
Chapter 5 Microstructure o(Hydrated Cement Pastes U8
occurs sporadically, and to a limited extent in portland cement concrete, and is more
commonly found in lime-pozzolan systems.
5-7-4. Carbonates
Deterioration of concrete service may be the result of a variety of physical and

chemical processes, such as attack by acids, sulphates, or alkalis, alkali-aggregate
reactions, freeze-thaw cycles, etc. These transport mechanisms include diffusion of
such gases as O2, CO2 or S02 in the gaseous phase of the pores, from the
environment to interior regions. Of the hydrates in the cement paste, the one which
reacts with CO2 most readily is Ca(OHh, the product of the reaction being CaC03 ,
but other hydrates are also decomposed, hydrated silica, alumina, and ferric oxide
being produced.
CaC03 dehydrates to calcium oxide ( CaO ) at about 900°C, with strong

endothermic peak on D.T.A. The phase of CaC03 ( calcite) gives five strongest x-
ray diffraction peaks corresponding to the 3.04 A, 2.29 A, 2.10 A, 1.91 A, and 2.50 A.
Carbonation shrinkage is probably caused by the dissolving of crystals of

Ca(0H)2 while under a compressive stress ( imposed by the drying shrinkage) and
depositing CaC03 in spaces from stress, the compressibility of the hydrated cement
paste is thus temporarily increased. If carbonation proceeds to the stage of
dehydration of C-S-H, this also produces carbonation shrinkage.
148
Chapter 6 Experimental Programme and Characteristics oU..,1aterials in Part II 149
CHAPTER SIX
EXPERIMENTAL PROGRAMME OF PART II
6-1. INTRODUCTION
The study of the microstructure mechanism is a great benefit in understanding the

properties of macrostructure in concrete science, and can predict the behaviour of
concrete macrostructure. There are many analytical procedures which can be used to
determine the microstructure mechanism of concrete such as X-ray diffraction,
mercury intrusion porosimetry, and scanning electron microscopy.
The effects of physical properties, mineralogical compositions and pore size

distribution due to pozzolanic activity in concrete containing high volume fly ash
were described in Chapter 4. The present study concentrates on the microstructure
development, and the mechanism of the fly ash reaction process occurring in different
proportions of high volume fly ash concrete combinations.
During this study, mercury intrusion porosimetry (MIP ) was introduced and
used to perform pore size distribution analysis to characterize the particle size of the
cementitious constituents of different mix proportions. X-ray diffraction ( XRD )
analysis technique is used to identify the mineralogical products, and monitor the
progress of hydration. This study also examines the effects of HVF A concrete on
permeability and water absorption, in order to investigate the density of HVF A
concrete under pozzolanic activity, and the effect of short and long term hydration
reaction. In this chapter, the experimental programme and the preparation of test
samples used in Part II of the research project are described.
149
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part 11 150
6-2. CONCRETE MIXTURES AND CHARACTERISTICS OF

MATERIAL
The material used and mixing proportions of the mixes were the same as those used
in Part I of the project, and have been discussed in Chapter 3, Section 3-3.
6-3. TEST PARAMETERS, PREPARATION OF TEST

SPECIMENS
In the present investigation, a series of concrete samples of the same concrete mix
proportions were tested at corresponding ages to correlate microstructure
development with strength characteristics. Immediately following compressive
strength testing of the specimens, experimental samples were taken from the centre of
each specimen after cutting away all surfaces to be free of the mold-wall impression.
The samples were then split into smaller pieces and immersed in high purity methanol
in order to avoid further hydration of cement, and dissolution of hydration products.
The samples were kept in methanol for at least 21 days before use. The
microstructure mechanism tests were carried out for all the concretes after 28, 90
days and 18 months of moist and air curing.
6-4. TEST DETAILS AND PROCEDURE
6-4-1. Mercury intrusion porosimetry study
The technique used to measure the pore size distribution was mercury intrusion
porosimetry. Mercury intrusion porosimetry is based on the capillary law governing
liquid penetration into small pores. The pressure required intrude the pore is a
function of the contact angle, the surface tension of the intruding liquid, and the
shape of the pore to be intruded.
150
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II J5J
A general expression, in the case of a non-wetting liquid like mercury relating

to these factors is given by the Washburn Equation
y<l> cose
D=- --------------- ( 6-1 )
P
where:
P = the pressures required to intrude into the pore
D = the narrow dimension of the pore
y = the surface tension of the liquid in dyne / cm
e = the contact angle between the mercury and the pore wall
<I> = the shape factor of the pore
The pore system of concrete is complex due to the morphology of the

hydration products being of "needle-like", "wall-like", or "amorphous structure".
Therefore, the feature of pore shape could be cylindrical, elliptical or slit pores. From
the surveyed literature for porosimetry on concrete materials, the shape factor ( <I> )
for the cylindrical and the slit pore is 4 and 2 respectively, while the elliptical pores
take the values between 2 and 4 [109].
Most researchers assume the pores to be of cylindrical cross section, so

selection of a shape factor ( <I> ) of 4 will yield results directly comparable to most of
the published research. Such a selection may not be the best to represent pores in
actual materials, but its use is generally accepted, as the practical means for treating
what would be a most complex problem.
The contact angle ( e ) between mercury and the solid containing the pores
varies somewhat with the methods adopted for drying the concrete sample, maturity
of the concrete and solid composition [110]. Winslow and Diamond [111] measured
the contact angles ( e ) for specimens oven dried at 105°C in oven and dried over
magnesium perchlorate hydrate. These measurements indicated contact angles ( e )
of 117 degrees and 130 degrees for the oven dried samples and perchlorate dried
respectively.
151
Chapler 6 Experimental Programme and Characteristics oOlaterials in Part 11
The volume ( V ) of mercury penetrating the pores is measured directl as a

function of applied pressures ( P ). This P-V information provides a set of pressure
and corresponding values of the cumulative pore volume From the pore size
distribution calculation, cumulative pore size distribution and differential pore
distribution ( dV I dD ) or dV/d ( log D ) can also be determined.
The pore size distribution studies were carried out using a Micromeritics
Poresizer ( Model 9320 ). A photograph of the instrument is shown in Fig 6- 1
Samples used for the porosity test were fragments of concrete by taking from the
internal parts of specimens with all coarse aggregate removed. These fragments were
then broken into pieces of approximately 3 rnm, and immersed in methanol for at
least 21 days. Prior to testing, the samples were oven-dried until they were of a
constant weight.
Figure 6-1 Photograph of the Micromeritics Poresizer ( Model 9320 )
About 1.5 - 2 grams of sample were then weighed, and inserted into a glass
penetrometer tube of known weight and volume. This penetrometer tube was then
fitted with a sealing cap, and a special grease ( Apiezon H ) applied to provide an aIr-
tight seal. The sealed penetrometer tube was then inserted into the low-pressure port
152
Chapter 6 Experimental Programme and Characteristics ofMaterials in Part II 153
of the Poresize 9320, and de-gassed by evacuating to a pressure below 50 llmHg, to

remove any remaining moisture and methanol.
After de-gassing, the penetrometer tube was filled with mercury, at a filling
pressure of 0.5 - l.5 psia, and the combined weight of the penetrometer, sample and
mercury was recorded. The penetrometer tube was then placed in the high pressure
port of the poresizer, where pressure was increased slowly from atmospheric up to
30,000 psia, and the amount of mercury intrusion measured at various points during
the pressure increase. A dedicated computer fitted with a maths co-processor was
used to control the instrument to record data, and to calculate various parameters.
The penetrometer used had a bulb volume of 5.89 mL and a stem volume of
1.131 mL. The value used for the surface tension of the mercury ( y ) was 485.00
dynes / cm. The shape factor ( <I> ) used was equal to 4, assuming all the pores to be
of cylindrical shape. For the pore diameter calculations a contact angle ( e ) of 117
degrees was used for the present samples, ( oven dried to 105°C to a constant
weight ), based on the previous investigation carried out by Winslow and Diamond.
Table 6-1 presents a typical output obtained with a pressure run from the present
study.
153
Table 6-1 A typical output of micromeritics pore size analyses
POROSITY OF CONCRETE SAMPLES
;PORESIZER 9320 V2.04 PAGE 1

SAMPLE DIRECTORY!NUMBER : HUNG 1022
OPERATOR: HUNG LP 11 :28:43 06/29/9t
SAMPLE ID : 350FW18M HP 11 :57:04 06/29/9t
SUBMITTER: REP 12:06:18 06/29/9t
PENETROMETER NUMBER: 10 ADVANCING CONTACT ANGLE: 117.0 deg.

PENETROMETER CONSTANT: 21.63 RECEDING CONTACT ANGLE: 117.0 deg.
pL/pF
PENETROMETER WEIGHT: 67.6639 g MERCURY SURFACE TENSION: 485.0
dynelem
STEM VOLUME: 1.1310 mL MERCURY DENSITY: 13.5512 glmL
MAXIMUM HEAD PRESSURE: 4.4500psi SAMPLE WEIGHT: 1.6654 g
PENETROMETER VOLUME: 5.8900 mL SAMPLE + PEN + Hg WEIGHT: 138.1900g
LOW PRESSURE :
MERCURY FILLING PRESSURE: 1.4230 psia
LAST LOW PRESSURE POINT: 1.4230 psia
HIGH PRESSURE :
RUN TYPE : AUTOMATIC
RUN METHOD: EQUILIBRATED
EQUILIBRATION TIME: 5 seconds
MAXIMUM INTRUSION VOLUME: 0.0010 mL/g
INTRUSION DATA SUMMARY
TOTAL INTRUSION VOLUME = 0.0348 mL/g

TOTAL PORE AREA = 14. 674sq-m/g
MEDIAN PORE DIAMETER ( VOLUME) = 0.0090pm
MEDIAN PORE DIAMETER ( AREA ) = 0.0073pm
AVERAGE PORE DIAMETER ( 4V/A) = 0.0095pm
BULK DENSITY = 2.0599 glmL
APPARENT (SKELETAL) DENSITY = 2.2189 glmL
POROSITY = 7.16%
STEM VOLUME USED = 5%
15.t
Chapter 6 Experimental Programme and Characteristics ofJvfaterials in Part II 155
Table 6-1 A typical output of micromeritics pore size analyses ( Contd.l )
POROSITY OF CONCRETE SAMPLES
PORESIZER 9320 V2.04 PAGE 2

SAMPLE DIRECTORYINUMBER : HUNG 1022
OPERATOR: HUNG LP 11 :28:43 06/29/9€
SAMPLE ill : 350FW18M HP 11 :57:04 06129/9~
SUBMITTER: REP 12:06:18 06/29/9€
PRESSURE PORE LOGDIFF ICUMULATIVE ~CREMENTAl DIFFEREN.

pSIa DIAMETER VOL dVIdlogD VOLUME VOLUME VOLdV/dD
pm mL/g mL/g mL/g mL/g-pm
0.49 1260.3412 9.67E-05 Kl P 1.58E-07
30.28 14.2183 15.78E-04 10.0002 P·0002 5.81E-05
50.78 12.5156 13.55E-03 10.0007 P.0005 15.99E-04
~9.78 1.8307 13. 19E-03 10.0012 P.0005 17.41E-04
101.44 1.2593 /2.57E-03 0.0017 10.0004 8.67E-04
149.44 10.8548 12.42E-03 10.0021 p.0004 1.21E-03
199.93 0.6389 12.40E-03 10.0024 p.0003 1.59E-03
~52.27 0.5064 /2.94E-03 0.0026 0.0002 12.46E-03
305.43 Kl.4182 13.43E-03 10.0029 p.0003 3.48E-03
1352.26 0.3626 13.39E-03 10·0031 p.0002 13.97E-03
~02.43 10.3174 13.28E-03 0.0033 P.0002 ~.36E-03
~55.09 0.2807 12. 62E-03 10.0035 0.0002 3.97E-03
522.76 0.2444 ~.l4E-03 p.0036 0.0001 3.72E-03
754.09 10.1694 ~.57E-03 p.004 0.0004 6.44E-03
1007.26 0.1268 ~.00E-03 P·0043 0.0003 1.00E-02
1256.42 0.1017 ~.61E-03 10.0046 0.0003 1.09E-02
1507.09 10.0848 ~.57E-03 10.0048 0.0002 1.29E-02
1754.25 p.0728 ~.27E-03 p.005 0.0002 1.33E-02
~029.75 10.0629 ~.72E-03 p.0051 0.0001 1. 84E-02
~507.75 10.0509 1.96E-03 p.0055 0.0004 1.63E-02
3003.42 0.0425 ~.33E-04 10.0055 P ~.32E-03
3505.25 ~.0364 6.07E-04 P.0055 0.0001 ~.09E-03
4000.42 0.0319 1.79E-04 P.0055 0 ~.38E-03
4999.76 P.0255 1. 15E-03 p.0056 10 1.92E-02
5998.92 0.0213 ~.43E-03 P.0058 0.0002 1.08E-01
kl988.26 0.0183 1.07E-02 10·0063 9·0005 ~.49E-Ol
~986.25 10.016 ~.03E-02 10·0068 10·0005 ~.40E-Ol
8524.58 10.015 ~.50E-02 P.0081 0.0013 7.09E-Ol
~008.25 P.0142 ~.82E-02 10.0084 10.0003 ~.43E-Ol
10006.07 10.0128 ~.94E-02 10.0098 p.0014 ~.79E-Ol
10959.9 P.0117 3.67E-02 10.0111 p.0012 1.34E+00
12181.39 0.0105 ~.25E-02 10.0124 10·0014 12. 13E+OO
12502.05 10.0102 ~.70E-02 10.0135 10·001 12.37E+00
155
Table 6-1 A typical output of micromeritics pore size analyses ( Contd.l )
POROSIlY OF CONCRETE SAMPLES
rORESIZER 9320 V2.04 PAGE]

SAMPLE DIRECTORY/NUMBER: HUNG /022
PPERATOR: HUNG LP 11:28:43 06/29/96
SAMPLEID: 350FW18M HP 11 :57:0406/29/96
SUBMITTER: REP 12:06: 18 06/29/96
PRESSURE PORE LOGDIFF ~UMULATIVI tINCREMENTAI DIFFEREN.

pSla DIAMETER VOL dV/dlogD VOLUME VOLUME VOLdV/dD
pm mL/g mL/g mL/g mL/g-pm
10959.9 ~.0117 ~.67E-02 ~.0111 10·0012 1. 34E+00
1218l.39 10.0105 ~.25E-02 p.0124 10·0014 12. 13E+OO
12502.05 10.0102 ~.70E-02 10.0135 10·001 12.37E+00
13352.54 10.0096 6.57E-02 ~.0151 0.0017 12.92E+00
13986.54 0.0091 17.02E-02 10.0164 10.0013 3.26E+00
14208.53 10.009 t7.13E-02 10.0176 10.0012 13.37E+00
14813.35 10.0086 7.28E-02 10.019 K>.0014 13.59E+00
14998.69 10.0085 1'7.29E-02 0.0193 10·0003 13.64E+00
15786.51 10.0081 17.29E-02 10.0205 0.0012 3.83E+00
16380.17 0.0078 7.22E-02 10·0217 K>.0012 13.94E+00
17122.67 10.0075 7.10E-02 10.0232 K>.0015 ~.04E+00
17493.83 0.0073 t7.04E-02 0.024 0.0008 ~.lOE+OO
1821l.32 10.007 6. 89E-02 0.0251 0.0012 4.17E+00
19004.98 10.0067 6.73E-02 K>.0264 0.0013 4.26E+00
19915.14 0.0064 6.75E-02 K>.028 0.0016 4.47E+00
~0880.47 10.0061 ~.61E-02 10.0292 0.0012 4.58E+00
~2227.63 0.0057 ~.80E-02 10.0303 0.0011 4.29E+00
~2487.12 0.0057 5.58E-02 10.0311 0.0008 4.18E+00
~3634.45 0.0054 ~.53E-02 10.0323 0.0012 S.55E+00
~4156.61 0.0053 ~.08E-02 10.0339 p.0016 S.28E+00
~4990.44 10.0051 S.50E-02 p.0341 K>.0002 ~.91E+00
~7472.96 0.0046 1.86E-02 p.0341 K> 1.70E+00
~8318.45 0.0045 l.40E-02 ~.0341 10 l.32E+00
29757.12 0.0043 1. 16E-02 K>.0348 0.0007 l.16E+00
~9975.62 0.0043 1. 16E-02 p.0348 0 l.16E+00
156
Chapter 6 l'.:xperimenfal Programme and Characteristics of.\Jareria/s in Pan 1I 157
6-4-2. Air permeability
At the specific age of the air permeability test, a set of three specimens were drilled
from the full thickness of the concrete to produce 50 mm diameter core samples
These cylindrical cores were then sliced at each end to produce samples of between
30 ~ 40 mm height for air permeability testing. The test apparatus was developed at
the University of Leeds [112] and a general view of the cells in use is shown in Fig 6-
2.
Figure 6-2 A general view of "The Leeds Permeability Cells"
The permeability cell consists of a sample holder, pressure gauge, gas supply
and flow meter at the downstream side. The cell was designed to test concrete
specimens of 50 mm diameter and 10 mm to 50 mm height. Due to the sensitivity of
oxygen permeability measurements to the presence of moisture content in the
concrete, after the prescribed period of curing, the specimens were oven dried at 105
°C to constant weight, until weight change was less than 0.1 percent over a 24 hour
period ( BS 1881 : part 5 ). The sample was placed in a rubber cylinder inside a steel
ring cylinder. Gas was forced to flow in the vertical direction through the sample
157
Flow rates were recorded by using a 1.7 mm diameter bubble flow meter at one bar
pressure when a steady state of flow occurred.
6-4-3. Water absorption
The water absorption test in the present study was according to BS 1881 : Part 122 :
1983 "Method for determination of water absorption". Similarly, a set of three
specimens were core-drilled from the full thickness of the concrete to produce
cylindrical samples of75 mm diameter, with a height of 100 mm. These samples were
then used for determining water absorption. The core samples were then oven dried
at 105°C for 72 + 2 h, removed from the oven, and cooled for 24 ± 0.5 h in a dry
airtight vessel. Each sample was weighed and immediately and completely immersed
in a tank with its longitudinal axis horizontal and at a depth such that there was 25 ±
5 mm of water over the top of the sample. The samples were kept immersed in the
water tank for 30 ± 0.5 min. Each sample was removed and shaken to remove the
bulk of the water and dried with a cloth until all free water was removed from the
surface. Each sample was then re-weighed.
The measured water absorption, Woos, of the sample was calculated as the
increase in mass resulting from immersion expressed as a percentage of the mass of

the dry specimen :
w abs
= WSSD -
W
WOD X 100% --------------- ( 6-2 )
OD
where WSSD and WOD represent the weight of the concrete sample in the SSD (
Saturated-surface-dry) and OD ( Oven-dry) states, respectively.
A correction factor according to the length of the specimen was obtained

from the curve shown in Fig 6-3 [113]. The product of this correction factor and the
measured absorption was the corrected absorption, this being the equivalent
absorption ofa core having a length of75 mm.
158
Chapter 6 Experimental Programme and Characteristics o[JJateria/s 111 Part 11 159
1.2 r-----:----:-----:--~---------------,
1.1 ·······;·········1·········,·······:··················.... ----._---_._---------;------------
~
1 , . . ,. . . . ;; ·r·r······ t------------:------------I------------1---
o ~ [
'0
~ 0.9 ---------t----- ! i : i . .
u
§
.+=1
o
u
~ 0.8 1: ... 1<::·1:::"; f
0.7 ····,I····L····J,4
. , , . , .
0.6 ---------~ ---_.. ----1 .. -. ------i- --------- -~ --~.
+ _. - -~ --~
- - -- -- - - --- --i. + •• - _ •• - --. L -- --_. - ---~ ------ ------~ ----------- -~- -------.. -- ~---.
~ . . . ~ ~~.
O. 5 L--L....-l---'-----L----'---'-----'------L---L--L--'---..L.-.1-...-JL-..!--1.-L--L----'-~__"____L_...L..._...l
!
30 40 50 60 70 80 90 100 llO 120 130 140 150
Length in mm of 75 mm dia. specimen
Figure 6-3 Correction factor related to the length of the specimen [113]
6-4-4. X-ray diffraction
Diffraction can be described as a coherent scattering phenomenon in which a large

number of atoms cooperate. The phenomenon occurs when radiation interacts with a
regular array of scattering points whose spacing is of the same magnitude as the
incident radiation. A crystal contains a regular array of atoms in three dimensions
with spacing in the range 1 ~ 5 A.
X-ray radiation has wavelengths in the same range and is scattered by atoms
through an interaction with the electrons. The diffraction phenomenon can be
visualised as a reflection by a set of parallel planes of atoms lying in the plane.
Diffraction occurs when the Bragg Equation [114] is satisfied:
A = 2 d SinS --------------- ( 6-3 )
Were:
159
Chapter 6 Experimental Programme and Characteristics o(Materials in Part II 160
A = wavelength of radiation
d = distance between the planes
9 = the angle of incidence
Diffraction will occur for the many different sets of atomic planes which can
occur in the crystal. A monochromatic X-ray beam will be diffracted at different
angles by the crystal structure of different minerals.
A diffraction pattern is derived by detecting diffracted X-ray beams with a

radiation detector as a random array of crystals is rotated in an incident beam of X-
rays. Thus, 29 is measured experimentally, which can be converted into "d" spacing
using the Bragg equation. A pattern consists of a series of diffraction peaks or
reflections whose positions and intensities are unique for each chemical compound
which forms crystals. It can thus be used for identification of that compound ; or for
quantitative analysis, since the intensity of a reflection is proportional to the weight
fraction of the compound in a mixture.
A well-crystallised material will have sharp diffraction peaks in considerable

numbers. Poorly crystalline compounds, where long-range symmetry is disturbed,
have fewer reflections which tend to be rather broad and diffuse.
In the present investigation, X-ray diffraction was only used to do qualitative

analysis because the systems involved complicated phases which are difficult to
analyse. The powder method of X-ray diffraction was used in the present study.
A sample, pre-treated with methanol for at least 21 days, was dried m

laboratory and crushed to powder using a pestle and then the powder was sieved to
pass the 150 micron sieves. Samples were examined using a Phillips Norelco X-ray
diffractometer using Cu ( Ka ) radiation and a scan rate of 2° / min, scan step 0.02°,
time / step 0.01 ° over the range 10° '" 70°. The Cu ( Ka ) radiation of wavelength
1.541838 A was used as the X-ray source. Anyone power pattern is characterized by
a set of line positions 28 and a set of relative line intensities, "f'.
160
The powder pattern X-ray data manual on cement minerals are according
with the Joint Committee on Powder Diffraction Standards ( JCPDS ) and the
Powder Diffraction File ( PDF ), and these were used to identify compounds, by
comparing known patterns from the manual with the test-result pattern until a
suitable match was obtained.
161
CHAPTER SEVEN
TEST RESULTS AND DISCUSSION OF PART II
7-1. INTRODUCTION
In Chapter four the mechanical behaviour of HVFA concretes was investigated in

order to attempt to find relations between these properties and the engineering
characteristics of the HVF A concretes. To understand the chemical and physical
processes governing mechanical behaviour, an extensive study was carried out in this
research, in which changes in mechanical properties of various concretes were
investigated in parallel with a study of their physical and chemical structure.
The goal of this chapter is to investigate the role of fly ash in HVF A concrete
III microstructure mechanisms due to pozzolanic reaction. Control concrete
specimens containing AS TM Type I cement with mineral admixture were also
investigated for comparison purposes.
The pore structure of HVFA concretes with different replacement mixtures

was studied, and the effect of other factors such as curing regime, age, and
cementitious material content were examined. A similar study, investigating the effect
on air permeability and water absorption of HVF A is reported. The relations of the
pore structure with strength, permeability, absorption are then discussed. A detailed
discussion dealing with the changes of morphological phase in different mixtures
under different curing regimes is also presented.
In the present investigation, samples were taken from the same concrete test
specimens, and were tested at corresponding ages to correlate microstructure
development with strength increase. Details of concrete mix proportions and testing
procedures have been described in Chapter three and Chapter six respectively.
162
7-2. PORE SIZE DISTRIBUTION ON NORMAL OPC AND HVFA

CONCRETE
Pore size distribution is a very important parameter that will indicate the change in
pore size under different conditions, since it will be affected by experimental
parameters such as degree of hydration, curing temperature, ageing, water /
cementitious ratios, and cement composition [70,71,115]. Pore size distribution in
combination with strength may give a more clear idea what mechanism controls the
strength when the experimental parameters are changed.
In this section a detailed analysis of the pore size distribution curves will be
presented, and on its basis, some conclusions will be drawn about the relationship of
both types of curing conditions on the pore structure of HVF A concretes. Some
relations between pore structure and mechanical properties will also be presented.
In the present investigation, plots were made of various experimental

parameters, primarily : ( a ), intruded volume as a function of pore diameter, and ( b
), dv / d ( log d ) as a function of pore diameter. The most useful parameter is
provided by the dv / d ( log d ) against pore diameter plots, from which the
"maximum continuous pore diameter" or the "threshold diameter" could be
determined.
7-2-1. Effect of different mixes and curing conditions on pore size

distribution
The results of Mercury Intrusion Porosimetry studies on HVF A concrete are

summarized in the data presented in the following series offigures, Figs 7-1, 7-2, 7-3
and 7-4. For purpose of comparison the result of the normal OPC concretes are also
included in the above figures.
Under both types curing conditions with different curing periods, the total
accessible pore volume and porosity of OPC and HVF A concrete are also shown in
Tables 7-1, 7-2, 7-3, and 7-4.
163
Chapter 7 Test Results and Discussion of Part II 16-1
0.081-----1------------r-------~
28days
,----
OJ) 0.06
~ 350C
S
j 350F
------;t(-
~ 0.04
350SF
.~ ~
1 '. ..............
~
350S
~ 0.02 ..............
".
'.
I O.~~
..•·· .. ··-·~·
..·-.... ·-~
..·...... ··~---·
.... ·~
.... --.. ~~~~
0.0025 0.05 10 300

Pore Diameter, micro-meter
0.08r------r------------.....,--------
90days
~0.06 350C
e
j 350F
~ 0.04 350SF
j ~
350S
~ 0.02
C)
I O~~~~~-~~~
0.0025 0.05
.. ·..
10 300
0.08 ,---------r--------------,--------~
0.07
'"" 540days
~0.06
e 350F
'-' 0.05 ----¥--
j:!Z 0.04 350SF
~
~ 350S
.~ 0.03
]
80.02
0.01
0.05 10 300
Pore Diameter, micl1>-meter
3
Figure 7-1 Cumulative pore volume curves for the 350 kg/m binder content
OPC and HVFA concrete ( Water curing)
164
Chapter 7 Test Results and Discu.<,sion of Part II 165
28days
OIl 0.06
~ ! 350C
S
! 350F
• ---7(--
.................
.........
....
I~~fl350S
~
..........
.................
10 300
0.08 I------I""---------------r-------~
90days
~0.06 350C
S
j 350F
---7(--
~ 0.04
350SF
.~...................•......
~
.~ -+-
-3 350S
~ 0.02 ...................
0.05 10 300
0.08 r - - - - - - - - - . - - - - - - - - - - - - - - - r - - - - - - - - - - - - .
0.07
""' 540days
SO.06
S 350F
'-' 0.05 ~
j 0.04
~
350SF
------4-
~ 350S
'.g 0.03
'3
30.02
om
OL-~~~~~~~~~~~~~~~==~------~~--~
0.0025 0.05 10 300
3
OPC and HVFA concrete (Air curing)
165
0.08 1 - - - - - -1 - - - - - - - - - - - - , - - - - - - - - - - -
28days
0.06
F
bO
::J
e !--~
~
450F
]
~ 0.04 I-t-
<I) i 450SF
I _____
.~ i
.5
='
I 4'~S
~ 0.02
0.08 r - - - - - - - - - r - - - - - - - - - - - - - , - - - - - - - - - - - - ,
90days
~0.06
450C
S
~
450F
] -t-
~ 0.04
450SF
.~
.5 4508
='
~ 0.02 --*---
0.08 r - - - - - - - - - , - - - - - - - - - - - - - - - , . . - - - - - - - - - - - - ,
0.07
,-.,
540days
~0.06
e 450F
'--' 0.05 ---t-
j 450SF
~ 0.04
~
----
450S
'i 0.03
'3
~ 0.02
om
0.05 10 300
3
OPC and HVFA concrete (Water curing)
166
Chapter 7 Test Results and Discussion o[Part JJ 167
0.08
.,
28days
OIl 0.06
~ 450C
e
0
450F
] --t---
~ 0.04
d)
;;.
'..\
\
I 450SF
j
;:) \; 450S
8 0 .02
\~ -----*"-
0.08 i - - - - - - - r - - - - - - - - - - - - - , - - - - - - - -
90days
~0.06 450C
e
.u 450F
] --t---
~ 0.04 450SF
d)
;;.
j
;:) 450S
8 0 .02
-~~
0.08 . - - - - - - - - - - - , - - - - - - - - - - - - - - - . . - - - - - - - - - - - - ,
0.07
,.-., 540days
5 0 .06
e 450F
'--' 0.05 -t--
0"
§ 4508F
~ 0.04
o 4508
;>
.~ 0.03
~ 0.02
0.01
0.05 10 300
3
OPC and HVFA concrete ( Air curing)
167
Table 7-1 Total accessible pore volume ofOPC and HVFA concrete-Water
curing
(Unit: mL/g)
Days 28 90 540
350C 0.0280 0.0381 ---

350F 0.0634 0.0642 0.0348
350SF 0.0617 0.0609 0.0473
350S 0.0616 0.0584 0.0470
450C 0.0237 0.0255 ---

450F 0.0648 0.0556 0.0394
450SF 0.0522 0.0483 0.0354
450S 0.0546 0.0608 0.0428
Table 7-2 Total accessible pore volume ofOPC and HVFA concrete-Air curing
(Unit: mL/g)
Days 28 90 540
350C 0.0334 0.0373 ---
350F 0.0569 0.0685 0.0619
350SF 0.0760 0.0635 0.0647
350S 0.0696 0.0646 0.0513
450C 0.0309 0.0353 ---
450F 0.0585 0.0604 0.0615
450SF 0.0643 0.0555 0.0687
450S 0.0708 0.0629 0.0581
168
Table 7-3 The porosity ofOPC and HVFA concrete-Water curing

(Unit: %)
Days 28 90 540
350C 6.53 7.35 ---

350F 13.65 12.82 7.16
350SF 12.45 11.75 10.20
350S 12.56 11.38 10.26
450C 5.36 6.00 ---

450F 13.48 11.63 8.86
450SF 11.12 10.08 6.86
450S 11.51 12.16 9.23
Table 7-4 The porosity ofOPC and HVFA concrete-Air curing

( Unit: % )
Days 28 90 540
350C 7.78 7.22 ---

350F 11.95 13.51 12.55
350SF 15.24 12.61 14.15
350S 13.37 12.75 11.03
450C 6.94 7.64 ---
450F 13.35 11.92 12.85
450SF 13.05 11.72 15.02
450S 14.48 12.78 11.75
169
As shown in the figures, total cumulative pore volumes are significantly

affected by the HVFA concretes. It was quite obvious from the figures, the total
volume of the geometrically continuous pores were higher for the HVF A concretes
than the normal OPC concrete with various curing periods. However, the pore size
distribution patterns appear to be different for the curing conditions and degree of
hydration with ages.
From the results, total cumulative pore size volumes of all concretes tend, as
expected, to decrease with increased hydration. At high degrees of hydration the
curves tend to become concave in the case of all concretes under water curing
condition. Meanwhile, the total accessible pore volume and porosity of OPC and
HVFA concrete toward to the lower value were found.
Under water curing, the concrete made with 350SF appeared to have more or
less the same cumulative pore volume of 350S, but the pore size distribution of this
porosity is quite different ; the former had most of its porosity in the coarse size
range, whereas that concrete of350S is mostly in the fine size range. These variation
of fine and coarse pores due to pozzolanic activity will be discussed in the next
section.
At the end of 90 days, for all concretes under water cunng, the total
cumulative pore volumes are lower than 28 days. For the long term pore size
distribution test at 540 days of all HVFA concretes under water curing, as expected,
the total cumulative volumes of the geometrically continuous pore were low than at
the ages of 28 days and 90 days. Meanwhile, the concrete made of 450SF had lowest
percentage in porosity when compared to all of the HVFA concretes. This indicates
that the amount of hydration products in 450SF were proportionately more than that
other HVFA concretes. The presence of the silica fume in the HVF A concrete was
thus instrumental in causing the beneficial effect observed on concrete strengths. This
also explained the reason why the crushing pattern of the specimen containing fly ash
or with both fly ash and silica fume was more ductile than that concrete without
above mineral admixture.
170
Chapter 7 Test Results and Discussion ofPart II 171
The results also show that the use of air curing regime was a remarkable
factor in determining the pore size distribution. These tends reflect a marked change
in the total cumulative pore volume ; meanwhile, the percentage of porosity are
increasing with continued air drying. Both the normal ope and HVF A concretes
showed higher accessible pore volumes compared to the corresponding water curing
concretes. Irrespective of normal ope concrete, at the age of 28 days, the concrete
made of350F had the total cumulative pore volumes lower than the other two HVFA
concretes, Mix-350SF and Mix-350S. The concrete made of mixes containing fly ash
as a partial replacement for both sand and cement, Mix-350S, Mix-450S had good
ability to resist the loss of the potential strength in the long term by air dried
condition.
From the results, at the end of 540 days the less the porosity of HVF A
concrete in water curing, the more is the value in the air regime. This is in agreement
with the preceding discussion in swelling and drying shrinkage of HVF A concretes.
The concretes made of 350SF, 450SF showed significant reduction in total pore
volume and, thus, had a good ability to resist swelling, but higher drying shrinkage
strain were also found on both types of concrete.
7-2-2. Maximum continuous pore diameter of pore structure
The maximum dv / d ( log d ) and the threshold diameter of ope and HVFA
concrete under both types curing conditions are shown in Tables 7-5 and 7-6. The
corresponding maximum dv / d ( log d ) against the threshold diameter on pore size
distribution are illustrated in the Figs 7-5, 7-6, 7-7, and 7-8.
From the results, the differential pore volume plots clearly demonstrated the
effect of HVFA concretes on the pore size distribution. At those ages under water
curing, although the total cumulative pore volumes of the HVF A concrete were much
higher than the normal ope concrete, the differential pore volume showed a higher
peak with a wider spread for the fine pores than that ope concrete. At the age of 28
days concrete made of 450S had much more pores concentrated at pore diameter
smaller than 0.023 pm in comparison with the ope concrete under water curing.
171
Under similar conditions, irrespective of normal OPC concretes, the concrete made of
450S had cumulative pore volume lower than the other HVFA concretes, meanwhile,
had most of its porosity spreading in the fine size range. The peaks of concrete mixed
with cement alone replaced partial by fly ash, 350F and 450F, occur at a significantly
larger pore diameter range than for the comparable concretes made with silica fume.
The maximum pore diameters of the concrete 350F, 350SF and 350S were at 0.0406
pm, 0.0286 pm, and 0.0244 pm respectively. Concrete made of 450F, 450SF, and
450S, the maximum continuous pore diameter were at 0.0481 pm, 0.0327 pm, and
0.023 pm respectively. From the results, however, the "maximum continuous pore
diameter" is a useful parameter in comprehending the concrete permeability. If
permeability is directly related to the maximum continuous pore diameter, then the
concretes of350S and 450S would be the least permeable.
At the end of 90 days, again, the total cumulative pore volumes also were
found to be lower in the concrete of350S, and had much more than the other HVFA
concretes. Surprisingly, concrete Mix-450SF had a lower cumulative pore volume,
meanwhile, gave highest differential pore volume for the fine pores and lowest
differential pore volumes for the coarse pores when compared to the normal OPC
and all HVF A concretes. This indicates the presence of many pores in the fine pore
range and a fewer number of pores in the coarse pore range for the concrete of
450SF in comparison with the concretes of 450S and 450F. Whereas, the concrete
made of 450S had higher cumulative pore volumes than the other concretes, the
differential pore volume is toward the bigger pore size distribution when compared to
the concrete of 450SF. The maximum continuous diameter of the concrete Mix-
450SF and Mix-450S were at 0.0107 pm and 0.0196pm. These variations in the
volume of fine pores in the concrete, although difficult to relate directly with concrete
properties, are nevertheless reflected in the strengths of corresponding concretes. The
addition of silica fume did, however, shift the distribution of pores toward the small
sizes. For the long term test at 540 days, the dv / d ( log d ) curves for both types
cementitious content concretes toward to the finer pore size distribution were found.
Although the concrete made of 350F and 450F had lowest total cumulative pore
volume when compared with the other HVFA concretes, the total coarse pores
172
volume had rather high. Reduction in the volume of large pore was observed with the
progress of the pozzolanic reaction. High concrete strengths were generally
associated with lower volume of large pores in the concrete. The distribution of pores
moved toward to the finer diameter and this explained the phenomena of gradual
increase in strength of concrete containing fly ash and silica fume, even though it had
higher total pore volume than concrete without fly ash or silica fume.
From the Figs 7-6 and 7-8, the different pore plots clearly demonstrated the
effect of air drying on the pore size distribution. All the concretes showed an increase
in number of coarse pores accompanied by a decrease in number of fine pores.
Irrespective of normal ope concretes, the concrete made of 350F and 350SF had
most of it porosity in the coarse size range compared to the concrete of 350S at the
age of 28 days. The maximum continuous pore diameter of the concretes 350F,
350SF, and 350S were at 0.0853 pm, 0.0718 pm, and 0.0229 pm respectively.
Although the total cumulative pore volume of the concrete 450S was much higher
than the concretes made of 450SF and 450F, the differential pore volume had shown
two peaks with a wider spread at the fine pores range. The maximum continuous
pore diameter were at 0.0573 pm for Mix-450F, 0.0553 pm for Mix-450SF, and
0.0531 pm, 0.0219 pm for Mix-450S respectively. At the end of 90 days, the
concrete made of 350F had a highest total cumulative pore volume and the
differential pore volume and showed the maximum continuous pore diameter at
0.0991pm , which is coarse pore range. The concrete 450S still had a lower
cumulative pore volume than that of concretes 450SF and 450F, but, the differential
pore volume had a peak concentrated at fine range. It is worth pointing out that the
concrete made of 450SF had a higher cumulative pore volume as well as the
differential pore volume had a peak in the coarse pores higher than 0.0751 pm, which
located in the coarse pore range. For the long term test on the air curing of HVF A
concretes, the dv / d ( log d ) curves for both types cementitious material content
concretes were found toward to the coarse pore size distribution. The above results
were found to be consistent with the discussion in engineering properties. In general,
engineering properties of concrete is influence by the volume of all voids in concrete,
capillary pores, gel pores, and entrained air, if present. From the result, it could be
173
inferred that, for hydration of concrete to continue, it is sufficient to prevent the loss
of moisture from the concrete.
Table 7-5 The max. dv/d ( log d ) and threshold diameter of OPC and HVFA
concrete-Water curing
Days 28 90 540
Mix Max. Threshold Max. Threshold Max. Threshold
dv/d-Iog d diameter dv/d-Iog d diameter dv/d-Iog d diameter
( mL/g-llffi ) ( Ilffi) (mL/g-j..I.ffi) ( Ilffi) ( mL/g-llffi ) ( Ilffi )
350C 0.0592 0.0129 0.0621 0.0141 --- ---
350F 0.0745 0.0406 0.0514 0.0223 0.0729 0.0081
350SF 0.0873 0.0286 0.0657 0.0206 *0.0706 0.0102
**0.0708 0.0073
350S 0.0913 0.0244 0.0914 0.0178 0.0866 0.0086
450C 0.0556 0.0162 0.0589 0.0129 --- ---
450F 0.0988 0.0481 0.0790 0.0200 0.0677 0.0069
450SF 0.0651 0.0327 0.0725 0.0107 0.0412 0.0094
450S 0.0881 0.0230 0.0898 0.0196 0.0702 0.0116
Table 7-6 The max. dv/d (log d) and threshold diameter ofOPC and HVFA
concrete-Air curing
Days 28 90 540
Mix Max. Threshold Max. Threshold Max. Threshold
dv/d-Iog d diameter dv/d-Iog d diameter dv/d-Iog d diameter
( mL/g-llffi ) ( Ilffi) ( mL/g-llffi ) ( Ilffi) (mL/g-llffi ) ( Ilffi)
350C 0.0594 0.0128 0.0383 0.0170 --- ---
350F 0.0596 0.0853 0.0669 0.0991 0.0536 0.0932
350SF 0.0522 0.0718 0.0440 0.0182 0.0603 0.1151
350S 0.0607 0.0229 *0.0501 0.0503 0.0621 0.0781
**0.0439 0.0156
450C 0.0352 0.0134 0.0363 0.0142 --- ---
450F 0.0582 0.0573 *0.0464 0.0751 0.0628 0.1160
**0.0500 0.0193
450SF 0.0506 0.0553 0.0451 0.0863 0.0679 0.1173
450S *0.0649 0.0531 0.0514 0.0169 0.0703 0.0650
**0.0567 0.0219
Remarks : * - Peak 1 and ** - Peak 2
174
0.1 i---------r------------------
2 0.08 28days
·sI
350C
0.06
350F
I!o
g ~
350SF
"e 004
co· ~
,g 350S
'-'
~
"e 0.02
0.1 r-------------~------------------------,-------------~
~ 0.08 90days
~
·s~ 0.06
350C
350F
i t ~
-*-
350SF
: i
!0.04 I : ~
,g
'-'
!\
I :
350S
~ 1
:
\:
"e 0.02 i 1
... . ".""." ::
~
I
..............
0.0025
. . .-.~.--. . ~
o~~~==~~d . -~~ 0.05 10 300
0.1 , - - - - - - - - - - - - - - , - - - - - - - - - - - - - - - - - - - - - - - - - - - , - - - - - - - - - - - - - - -
0.09
~ 0.08
540days
¥0.07 350F
~ 0.06 -*-
Jio 350SF
80.05 ------4-
"e 0.04 350S
~
'-' 0.03
"e
> 0.02
"0
om
P<Te Diameter, micro-meter

J
Figure 7-5 Differential pore volume curves for the 350 kg!m binder content
ope and HVFA concrete ( Water curing)
175
0.1 ,-----1-----------,-------
... 0.08 28days

~
·s~ 0.06
350C
350F
t!o
'8 -----*-
350SF
-0 004 ~
~.
350S
~
-0 0.02 \
o~~~~~··--~-----=-----=---=----=----=---=----=---=---~-----=---=--~----~---~-~~~~~~
, 0.0025 0.05 10 300
0.1 1 - - - - - - " 1 " " - - - - - - - - - - - , - - - - - - - -
90days
.s 0.08
(I)
350C
·s~ 0.06 350F

-in -----*-
8 350SF
~
!0.04
.Q 350S
'-'
~
-0 0.02
0.05 10 300
0.1 r--------------,-------------------------,--------------~
0.09
1:3 0.08 540days

"0
30.07 350F
8
S 0.06 ~
i 0.05
3508F
----4--
3508
-0 0.04
]>
'-' 0.03
-0
~
-0 0.02
0.01
o~~~--~~~~~~~~==~~~~====----~
0.05 10 300
0.0025
3
Figure 7-6 Differential pore volume curves for the 350 kg/m binder content
ope and HVFA concrete ( Air curing)
176
0.1 I-----Tr-----------,-------~
... 0.08 28days

~
El
g
·s 0.06
F- - ,
! 450F
~
"8 i-+-
I
,..::. I 450SF •
4~S
"0 0.04
tlll
.2
"0
I ~:
:
> 0.02
"0
0.1r-~----------_,--------------------------,---------------~
90days
~ 0.08
450C
~...
·s 0.06
()
450F
~
0() -+-
,..::.
450SF
~ 0.04
.2 450S
......,
~
"0
>
"0 0.02
0.1 ,--------------,--------------------------,----------------,
0.09
a 0.08 540days
o
~ 0.07
450F
~ 0.06 -+-
-in 4508F
8 0.05
"0 0.04 4508

~
......, 0.03
"0
> 0.02
"0
om
0.05 10 300
3
Figure 7-7 Differential pore volume curves for the 450 kglm binder content
ope and HVFA concrete (Water curing)
177
Chapter 7 Test Results and DiscussIOn uf Part II 178
O.l,-----i----------------_---,
~ 0.08 28days
"'Q)
S 450C
g
·s 0.06
- - I
450F
--tl.o
()
() -t--
450SF
I
~
-.;)
1;)/)
004
•
--
.sa 450S
.....-
-.;)
;.
-.;) 0.02
0.1 r--------------r-------------------------,---------------
~ 0.08 90days
"'Q)
S 450C
§
.~ 0.06 450F
--
I;)/)
() -t--
()
~
450SF
-.;)I;)/) 0.04
.sa
-.;)
;.
-.;) 0.02
--
450S
0.1 .--------------,---------------------------,-----------------,
0.09
~ 0.08
540days
¥... 0.07
450F
~ 0.06 -+-
~ 450SF
8 0.05
-.;) 0.04 450S

~
......, 0.03
-.;)
;.
-.;) 0.02
0.05 10 300
3
Figure 7-8 Differential pore volume curves for the 450 kglm binder content
ope and HVFA concrete (Air curing)
178
7-2-3. Comparison with the total binder content on pore size

distribution
In comparison with total binder content on all of concretes under both types of curing
conditions are shown from Fig 7-9 to Fig 7-20. The results show that the total binder
content does not seem to have much effect on the pore size distribution. Concretes
3
made with 450 kg/m mixes were slightly lower cumulative pore volumes than that
3
concretes 350 kg/m mixes under both types curing conditions. It is importan
pointing out that, although concrete made of 450S had higher cumulative pore
volumes, the pore size distribution had similar trend 350S under water curing at 90
days. The maximum continuous pore diameter of both types of concrete approach to
fine pore size range, which is 0.0178 Ilm for Mix-350S, 0.0196 Ilm for Mix-450S
respectively. Similar results were found in air curing; the maximum continuous pore
diameter of Mix-450S had a finer pore size distribution at second peak than that of
concrete of350S.
It is interesting in the differential pore size plots, the concrete of both types of
binder content under the same curing condition and age, had more or less a similar
tendency in differential pore size curve. Porosity is a major factor in controlling
strength and modulus of elasticity of concrete systems. As demonstrated in preceding
discussion on engineering properties, the pulse velocity is controlled primarily by the
dynamic modulus of elasticity of the concrete. Test concretes from all the mixtures
with both types cementitious material contents had shown similar behaviours of pulse
velocity under the same conditions. From the results, there is no doubt that the pulse
velocity of concrete is also influenced by the volume of all pores in concrete.
179
,-.. 0.08 ~ 008 r----,-------;----.,

~ 00·
'§
'-" 0.06
~'-" 0.06
g
~
Q)
~
E
~
, 350F ;
~ 0.04
, *
450C ~ 0.04
i 450F ,
~
.~ 0.02
~ ~
-"'\. .~ 0.02
1 \
~ .. t::-~
"3
~
E
u 0?0025 0.05 10 300 u 0~0~2:--::5-~0..J..0-5~.........I~="'"'10=--3WOO
Diameter, micro-meter Diameter, micro-meter
.....
$ 0.1 2o 0.1 .------::j.--------,---~
~ E
e0.08
o
e0.08
·s o
·s
~~J
~0.06 ~0.06 350F
8
~O.04 \!
450C 8 *
450F
,
I
I
~0.04 +
t .i
\
Q[) Q[)
..9 0.02 I,'
\.~ ..9 0.02
~
,..,._.
0~025 0.05
-
"0 10 300 0.05 10 300

( Water curing 28 day)
,-.. 0.08 ,----,----------.-----, ~0.08
~'-" 0.06 ~'-" 0.06

0' 0'
350SF E 350S
§ ~ ~ -B-
~ 0.04 450SF ~ 0.04 450S
-.
~ ~
.~ 0.02 .~ 0.02
"3
1 E
~
u o~--~~~~~k==-~
u 0~025 0.05 10 300 0.0025 0.05 10 300
.....
2 O.l,----,----------,-----, ~ 0.1 ..------~----...,------,
8 E
e
o
0.08 e0.08
o
·s ·s 350S
~0.06 350SF ~0.06 -B-
8 ~
450SF '8 450S
~0.04 ~ 0.04 -.
"0 "0
tl() tl()
..9 0.02 ..9 0.02
~
"0
~
"0 0~025 0.05 10 300 "0 0~025 0.05 10 300
Figure 7-10 Cumulative and differential pore volume of 350SF/450SF,

350S/450S (Water curing 28 day)
180
Chapter 7 Test Results and Discussion of Part II 181
""'
bIl
0.08 ~ 0.08 r-----,-------,---~
~'-../ 0.06 ~'-' 0.06
1,;) I:)~
S
;:l
~ 0.04
~
... @J
450C
S
;:l
~ 0.04
~
350F
-r-
450F
+
.~
1
u
0.02
0?0025
\::\."'. -
0.05
.... .........•.
.~
"3
U
S
;:l
0.02
O?oO'-2~5-~0......0-5~-~~....l10-*--3-t.l
10 300 OO
2 0.1 ~
... 0.1 , - - - - , - - - - - - - , - - - - - ,
S S
8o 0.08 8o 0.08
's~0.06 '8
~
~0.06 350F
8 450C 8 *
450F
""' 0.04
'"C)
""' 0.04
'"C)
+
00
.Q 0.02 (; 00
.9 0.02
.... . .
-
0.05 10 300 0.05 10 300
Figure 7-11 Cumulative and differential pore volume 350C/450C, 350F/450F

( Water curing 90 day)
,-... 0.08 , - - - , . - - - - - - - r - - - - - - , ""' 0 .08 , - - - - - - r - - - - - - - , - - - - - ,

~ 00
]
'-' 0.06
~'-' 0.06
~
I:)~
350SF 350S
;:l
~
§ -&
~ 0.04
~
~
0.04 450SF
~
...
450S
.~ 0.02 '.g 0.02

1 ];:l
u O?o025 0.05 10 300 u O?o025 0.05 10 300

I 0.1
i
S
0.1 , . - - - , - - - - - - - , - - - - - - ,
8o 0.08 8o 0.08
's 350SF
'8 350S
~0.06 00 0.06 -&
oo ~ ----8
~
'"C)
0.04
450SF
""' 0.04
'"C)
...
450S
bIl bIl
.9 0.02 .9 0.02
'"C)
::2
> 10 300
'"C) O?o025 0.05 10 300 .; O?o025 0.05

181
Chapter 7 Test Results and Discussion ofPart 11
182
'; 0.08 , - - - , - - - - - . , - - - - - - .
~'"-' 0.06
~~ 450F
~ 0.04
*
450F
~ +
.~ 0.02
'"3
S
8 0~OL2~5~~0~.0~5~~~~.....110~~-I3JOO
~ 0.1 r - - - , - - - - - - - - r - - - - - - ,
S
2
Q
0.08
·s
On 0.06 350F
'0
Q
r:. 0.04
*
450F
"'0
+
01)
...9 0.02
0.05 10 300

(Water curing 540 day)
,-... 0.08 , - - - - . , . - - - - - - - r - - - - - ,
~'--' 0.06
<If
350SF 350S
~
450SF ~::> 0.04 -&
450S
-.
~
~ 0.02
~
0.05 10 300 u 0~025 0.05 10 300
Pore Diameter, micro-meter Pore Diameter, micro-meter
~ 0.1,----.,.--------.-------, 1-0 0 ,--_ _,...--_ _ _ _, -_ _--,

~ .1
~I
~I
8 0.08 8u 0.08
.§ ·s00 0.06
.io 0.06 350SF
~
-...
350S
-&
8 450SF 8 450S
A 0.04 A 0.04 -.
'"C '"Cl
01) 01)
...9 0.02 ...9 0.02
~ ~
'"Cl 0~025 0.05 10 300 '"C 0~025 0.05 10 300

182
~ 0.08 r---,------......,-----, ,-., 008 r - - - - , - - - - - , - - - - - .

~.
I:l{)
~
'-' 0.06 '-' 0.06
S~
;::l
"00.04 ~ ~~
~ 0.04
350F
+
~
~
.~
~ ~
450F
+
0.02 .~ 0.02
~ ~
§ E
;::l
u O~~~~--------~~=-~
u 0?0025 0.05 10 300 0.0025 0.05 10 300
2 0.1 r-------,;------..,-----,
....
~ 0.1 r---,-------,---~
S E
e0.08
o e0.08
o
·s ·s
i O 06
. ~0.06
8
350F
*
450F
~0.04 ~0.04 +
'"..9 0.02
0/) '"
0/)
..9 0.02
~
'" 0?0025 0.05 10 300 0.05 10 300

(Air curing 28 day)
~0.08 ~ 0.08 ,.----,-------,---~

0/)
~'-..' 0.06 ~'-' 0.06

~~ 350SF
s~
350S
;::l
~ -B-
~ 0.04 450SF "0 0.04 450S
~
-*"
~ ~
.~ 0.02 .~ 0.02
"3 ~
§ §
u o~--~~--~--~~~--~
0.0025 0.05 10 300 u 0~025 0.05 10 300
....
2 0.1 ,.----.,.....-----......,.-----, ~ 0.1
S E
e0.08
o
e0.08
o
'seo 0.06 350SF
'seo 0.06 350S
....... ~
8 ~
450SF 8 450S
:
~ 0.04 ~0.04 -*" ;
0/) '"
0/)
..9 0.02 ..9 0.02

=2
~
'" 0~025 0.05 10 3 00 .; 0~025 0.05 10 300

350S/450S (Air curing 28 day)
183
Chapter 7 Test Results and Discussion o(Part 11
184
,-, 0.08 r - - - - , - - - - - - , . - - - - - , ~ 0.08 r---r------,---~

01)
~ ~ :K
'-' 0.06 '-' 0.06
Q)' Q)'
S S i 350F
::I ::I
"00.04 ~ 0.04
~ i 4;F
1;: 1;: +
.~ 0.02 .~ 0.02
"3 "3
§ §
u 0?0025 0.05 10 300 u 0 ?oOL2~5~~O...l.0-5--":!::::::::~,,,"*,J..l0-~4"3..J
OO
....
2 O.l r - - - - - - , - - - - - - , - - - - - , Q) 0.1 , - - - , - - - - - - , - - - - - ,
~
Q)
S
8 0.08
()
8 0.08
·s
()
'8
~0.06 ~0.06 350F
8
~0.04
8
~0.04
*
450F
+
-0 -0
01) 01)
.Q 0.02 .Q 0.02
~ ::2
-0 0~025 0.05 10 300 ~ 0~OL2~5-~0....L.0-5~+-~~.J.I0-~~300
Figure 7-17 Cumulative and differential pore volume of 350C/450C,

350F/450F (Air curing 90 day)
,-., 0 08 r - - - - - - , , - - - - - - , . - - - - - - , ,-., 0.08 r - - - - , - - - - - - - - , - - - - - ,

~.
01)
'-' 0.06
~'-' 0.06
ef 350SF ~' 350S
::I ~ -&
~ 0.04
1;:
450SF "0 0.04
~
1;:
...
450S
.~ 0.02 .~ 0.02
"'3
§ ~
u 0~025
8 0 L~---.l_=::::::~~==--..J
0.05 10 300 0.0025 0.05 10 300
.... ....
~ O.l,----.,--------.------, ~ 0.1
S S
8 0.08
()
8 0.08
()
's '8
~0.06
--800 0.06
350SF
~
8 450SF
~0.04 ~0.04
01) 01)
.Q 0.02 .Q 0.02
-0 ::2
> ~ 0~025
-0 0~025 0.05 10 300 0.05 10 300

184
('hapler 7 Test Results and Discussion of Part 11 185
,-., 008 r - - - - , - - - - - r - - - -
tll)'
~'--' 0.06
~~ 350F
~ 0.04 *
450F
~ +
~ 0.02
~ 0~025 0.05 10 300

.B 0.1 r---r------...,-----
SI
8u 0.08
·s
~0.06 350F
u
r: 0.04
*+
450F
"0
tll)
.Q 0.02
'-'
~
~ 0~025 0.05 10 300

(Air curing 540 day)
~ 0.08 , - - - , - - - - - - r - - - - - , ~ 0.08 , - - - - - , - - - - - - - , - - -
~ ~
~ ~
'-' 0.06 '-' 0.06
~"
/j)"
350SF 350S
~::> 0.04
~ 0.04
~
~
450SF
~
..
-e-
450S
! 0.02 j 0.02
~ ~
u 0~025 0.05 10 300 u 0~025 0.05 10 300
jI-<
0.1 ~
I-< 0 .1 ,-----r------,-----,
S
8 0.08 I
80.08
.§ .§
-inu 0.06 350SF do 0.06 350S
u
~0.04
'\j
tll)
~
450SF '8
r: 0.04
"0
tll)
..
-e-
450S
.Q 0.02 ..9 0.02

'-' '-'
'\j
">
'\j 0.~025 0.05 10 300 0.05 10 300

185
7-2-4. Total, macro-pore and meso-pore volumes
As discussed in the preceding section, the pore structure will be divided into three
parts following closely the division defined by IUPAC : macro-pores ( d > 500 A ),
meso-pores ( 25 A", 500 A ), and micro-pores ( d < 25 A ). The effect of different
curing conditions on the total macro-pores, and meso-pores distribution of both
types binder contents are presented in Tables 7-7 and 7-8.
The results show that all the concrete mixtures with continuous water curing
exhibited a higher percentage meso-pores than the air drying concretes. Meanwhile,
the use of air curing of all concretes a slight reduction in meso-pore, and macro-pore
are just relatively increase with continued air drying.
Under water curing, however, the HVFA concretes had much greater meso-
pores, when compared to the normal OPC concrete at the ages of 28 days and 90
days under the same conditions. The trends in the macro-pores are just the opposite.
The larger percentage of macro-pores of normal OPC concrete reflects its "coarser"
pore structure. It is reasonable to suppose that increase in meso-pores with increased
curing age is the result of growth of hydration products during hydration reaction. On
this table, it can be clearly seen that the concrete of 450SF the meso-pores range and
much less in the macro-pores range at the age of 90 days. Meanwhile, at the end of
540 days, the concrete made of 350S and 450S had a highest percentage of meso-
pores compared to the other HVFA concretes.
Macro-pore into meso-pore due to reduction of their pore size by the growth
of hydration products into the available spaces between the unhydrated grains, which
were initially filled with water [71].
Data from the results also showed that an increase in the accessible pore
volume in the presence of concrete with air curing is associated with a decrease in the
number of fine pores, and an increase in the number of coarse pores when compared
to the wet specimens at the same conditions. This shift in pore structure is exhibited
in the differential pore volume curves as were already shown in above Figs 7-2, 7-4,
7-6, and 7-8.
186
Chapter 7 Test Results and Discussion of Part II
Table 7-7 The macropore and mesopore on ope and HVFA concrete-Water
curing
( Unit: % )
Range > 0.05 J-lffi 0.05 J-lffi '" 0.0025 J-lffi

Days 28 90 540 28 90 540
350C 22.86 28.78 --- 77.14 71.22 ---
350F 23.66 20.56 15.80 76.34 79.44 84.20
3508F 16.21 15.27 13.53 83.79 84.73 86.47
3508 15.05 13.01 11.06 84.95 86.99 88.94
450C 15.08 25.10 --- 86.92 74.90 ---
450F 21.45 16.37 17.26 78.55 83.63 82.74
4508F 34.29 8.07 15.31 65.71 91.93 84.69
4508 13.92 16.86 13.32 86.08 83.14 86.68
Table 7-8 The macropore and mesopore on ope and HVFA concrete-Air
curing
(Unit: %)
Range > 0.05 J-lffi 0.05 J-lffi '" 0.0025 J-lffi
Days 28 90 540 28 90 540
350C 24.55 38.61 --- 75.45 61.39 ---

350F 61.16 65.40 78.03 38.84 34.60 21.97
3508F 62.11 43.78 67.85 37.89 56.22 32.15
3508 37.93 31.26 61.21 62.07 68.74 38.79
450C 23.62 26.63 --- 76.38 73.37 ---

450F 51.45 42.38 71.87 48.55 57.62 28.13
4508F 62.05 60.36 67.69 37.95 39.64 32.31
4508 47.60 36.41 61.03 52.40 63.59 38.97
187
The pore size distribution in the presence of air curing concrete mixtures
changed from a narrower to a wider distribution with a shift in meso-pore diameter
toward to a macro-pore diameter. However, the variation in macro-pore, meso-pore
indicated the air drying has a marked effect on the arrangement of these growing
hydration products.
7-2-5. Some relations between the pore structure and compressive

strength
The relationship between the compressive strength, fe, and porosity of concrete, X, is
by the power equation :
b
fe = a X --------------- ( 7-1 )
where fe is the compressive strength in MPa, X is the porosity in %, and a and bare
constants.
Although both mixes with 350 kg/m3 and 450 kglm3 total binder content
concretes had shown similar pore size distribution under the same conditions, the
curing regime was the marked factor in determining the result of compressive
strength, and the pore size distribution. In general, the water cured concretes showed
higher values of compressive strength than those exposed to drying concretes ; it is
suggested that separate equations be developed for each curing condition.
The relation between porosity and compressive strength for all concretes
under different curing conditions is shown in Figs 7-21 and 7-22. The relationship is
based on the data at 28 days, 90 days, and 18 months for all concretes under water
and drying curing conditions. Regression analysis of the data shown in the figures
produced the following relationships :
fe= 243.7 X -0.67 (moist) (r = - 0.75 ) --------------- ( 7-2 )
fc = 505.8 X -0.95 (dry ) ( r = - 0.72 ) --------------- ( 7-3 )
188
120
350C 350F 350SF 350S
110
o X o o
100 450C 450F 450SF 450S
~
~
90
• + •
t53 80
~
00 70
Q)
• • o
.....> +
rIl
rIl
Q)
60
~0
50
u
40
30
20
4 6 8 10 12 14 16
Porosity, %
Figure 7-21 Relationship between compressive strength and porosity of OPC

and HVFA concrete-Water curing
120
350C 350F 350SF 350S
110 o o
D X
100 450C 450F 450SF 450S
~
~
90
• + •
t53 80
~
00 70
Q)
.....>
rIl
~
60
a
§ 50
u
40
30 x + o
20
4 6 8 10 12 14 16
Porosity, 0/0
Figure 7-22 Relationship between compressive strength and porosity of OPC

and HVFA concrete-Air curing
189
Chapter 7 Test Results and Discussion o(Part II 190
The correlation factor with the two equations is low. These data indicate a
significant correlation between porosity and compressive strength in concretes at the
95 % confidence level. A single equation describing the two parameters can be used
as given below, and is shown in Fig 7-23.
fc= 288.4 X -0.74 (moist + dry) (r = - 0.71 ) ----------- (7-4 )
Again, the correlation factor between both parameters of equation is still low
and the equation relating the two is approximately the same for the water and air
drying curing concretes. All concretes have this same inverse dependency between
strength and porosity. Many workers [91,116] have made attempts to correlate
various mechanical properties with parameters related to pore structure, especially
total porosity. In fact, the effect of the degree of hydration on macro-pore and meso-
pore diameter obtained by Mercury Intrusion Porosimetry is shown in Tables 7-7 and
7-8. In most cases, very large changes occur in the diameter of the macro-pores and
meso-pores. The variation in macro-pores and meso-pores indicate that curing
condition and the degree of hydration have a marked effect on the arrangement of
these growing hydration products [70,71]. Generally, at a given porosity, smaller
pores lead a higher strength of the concrete. Regression analysis in Fig 7-24 shows
that the coefficient of determination ( r = 0.90 ), using meso-pore data in the
relationship, is very high. The results infer that high strength / performance concretes
are seen to be insensitive to the structure of the smaller pores, and are not to be
associated with the total porosity. As can seen in the preceding Tables 7-3, 7-4,
porosity over the whole size rang of pores has influence on concrete properties, yet it
is difficult to get an exact assessment of pore size distribution because Mercury
Intrusion Porosimetry measurement encompasses the whole size range and because it
is difficult to interpret experimental data. Thus, comparisons of porosity should be
made with care.
190
120
350C 350F 350SF 350S
no 0 X 0 0
100 450C 450F 450SF 450S
~
d
90
• + •
o.
tas 80
.p
••
rn 70 0
•
.-Q)
;..
til
til
Q)
60
0
0 0
0
,• •
0
'P-"' o
§ 50 ( MOIST + DRY)
U
40
0
30 x +-t 0
20
4 6 8 10 12 14 16
Porosity, %
Figure 7-23 Relationship between compressive strength and porosity of ope
and HVFA concrete
80
350F 350SF 350S 450F 450SF 450S
X 0 0 + •
d 70
~ •
tas
~ 60
-
Q)
;..
.til
til
Q)
0
•
~ Y = 25.996 X
" 0.20
(r = 0.90) (MOIST + DRY)
8 50
+
40 L-~~~~~ __L-~-L~~__~~~~~~--~
10 20 30 40 50 60 70 80 90 100
Meso-pore, %
Figure 7-24 Relationship between compressive strength and meso-pore of
HVFA concrete
191
7-3. PERMEABILITY OF HVFA CONCRETE
Permeability is one of the basic characteristics of concrete relating to its durability,

and especially when concerned with its use in special applications such as in the
isolation of water materials. If a concrete lacked impermeability, even though it
theoretically has a high strength, it would not be acceptable in most cases since its
durability would be at risk [94,117]. Moreover, permeability properties give
considerable insight into the mechanisms of fluid and ionic transport through
concrete. Relationships have been shown to exist between the air permeability and
carbonation resistance and the initial surface absorption and abrasion resistance of
concrete [118].
The permeability test is based on "The Leeds Permeability Cell" [112], which
measures the coefficient of permeability by measurement of through flow of oxygen.
Permeability tests were performed on 50 mm diameter 30 ~ 40 mm cores samples
drilled from 100 x 100 x 500 mm prisms. The test specimens were subjected to
different curing and tested at the age of 28 days, 90 days, and 18 months. For each
curing three cored samples were used, the results are thus the average of three
readings. Mix proportions and test procedure described in section 6-4-2. Variations
in different replacement methods and total cementitious content also are investigated.
The relationship between the compressive strength and permeability are also
reported. A relation between the air permeability and porosity for HVF A concretes
was found.
7-3-1. Effect of curing condition on air permeability
Under water curing with various curing periods, the data on the air permeability of
HVFA concretes with both types of binder contents are given in Table 7-9 and
illustrated in Figs 7-25 and 7-26.
192
Table 7-9 High volume fly ash concrete permeability coefficient

(Unit : 10-16 m2 )
Curimg Water curing Air curing
Days 28 90 180 540 28 90 540
350F 0.305 0.151 0.132 0.203 0.576 0.431 0.846
3508F 0.142 0.134 0.130 0.176 0.255 0.206 0.761
3508 0.126 0.127 0.121 0.163 0.234 0.160 0.454
450F 0.277 0.135 0.121 0.144 0.524 0.377 0.701
4508F 0.132 0.134 0.111 0.181 0.247 0.189 0.512
4508 0.117 0.102 0.101 0.194 0.217 0.160 0.437
Data from the results indicated that the coefficient of permeability was low in
water cured specimens at early ages for all HVFA concretes. Between the ages of 28
days and six months, obviously, the presence of silica fume in mixes 3508F and
4508F helped mixes 350F and 450F to develop a denser internal structure and a
lower permeability. Mixes 3508 and 4508 with large amounts of fly ash developed a
much denser structure and low permeability from early ages. As already stated in
section 5-5, the permeability of concrete is not only a simple function of its porosity,
but depends also on the size, distribution, shape tortuosity, and continuity of the
pores. The present investigation, from the preceding results of the pore size
distribution in Table 7-7, suggested that the permeability is strongly controlled by an
interconnecting network of capillary pores. At 28 days and 90 days under water
curing, the concrete made with 4508 had much lower macro-pores, compared to the
same condition of concrete 450F. The reduction in amount of macro-pores with
duration of specimens under water is correspondingly reflected in the permeability of
concrete.
193
Chapter 7 Test Results and Discussion orpar! II j;-/.
0.35 r--------------------~
350F 350SF 350S

~
)( 0 0
Ne
'-"
-......
'C)
b
0.3
\
....r 0.25
5
......
(,)
~
8
u 0.2
C
.....
......
~~ 0.l5 \\~
~ :~---~
0.1 ~ ____ ~ __ L_~~~~_ _ _ _ _ _ _ _ L __ _ _ _~_ _~_ _~~
28 90 180 540
Age (Days)
Figure 7-25 The permeability coefficient ofHVFA concrete-water curing

( 350 kglm3 mixes)
0.3 , - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
450F 450SF 450S
• •
Ne 0.25
-sn 0.2
.....o
(,)
~
o
o
u 0.15
C
.....
......
.....
'io
§
o
O.l
~
O. 05 LJ._ _~_L_~~~J.....__ _ _ _L __ _~_ ___'__...L_'____'
28 90 180 540
Age (Days)
Figure 7-26 The permeability coefficient of HVFA concrete-water curing

( 450 kg/m3 mixes)
194
Chapter 7 Test Results and Discussion o[Part II 195
It is clear from those results that the permeability of the HVF A concrete was
directly related to the pozzolanic reaction of fly ash and silica fume in concrete,
which produces calcium silicate hydrate, tends to fill these unoccupied spaces to form
a product with decreased permeability and which also results in improved strength
[28]. These results were also found by Manmohan and Methta [37]. They showed
that additions of pozzolanic and cementitious admixtures such as fly ash to portland
cement were instrumental in causing pore refmement or transformation of large pores
into fine pores - a process which had a far reaching influence on the permeability of
the hardened cement concretes. Meanwhile, the use of fly ash to replace both cement
and sand partially might have a good combination effect on permeability.
In general terms, it is possible to say that the higher the strength of the
hardened cement concrete the lower its permeability - a state of affairs to be expected
because strength is a function of the relative volume of gel in the space available to it
[69]. Nevertheless, from the figures, the improvement in strength with ages of water
curing seems not to be correspondingly reflected in the permeability of concrete. For
the long term under water curing, the concretes exhibited slightly increasing
coefficient of permeability at the age of 18 months when compared to the
permeability with the same conditions at the age of six months. In comparing both
porosity ( Tables 7-7 and 7-8 ) and coefficient of permeability of concrete at the age
of 504 days, the concrete made of 450F has a higher macro-pore volume but a lower
coefficient of permeability. Whereas, for concrete made of 450S a lower percentage
in macro-pores was found. The corresponding coefficient of permeability for the
concrete made of 450S were high. It would be expected, as a first approximation,
that a sample with a higher porosity also would have a large permeability. However,
this generality does not take into consideration the possibility of and potential effects
of differences in pore size distribution. Research data [9] indicate the sensitivity of
high strength fly ash concrete to curing, especially for the most mature concrete. In a
mature concrete, the permeability depends on size, shape, and concentration of the
gel particles and on whether or not the capillaries have become discontinuous.
In the present investigation, the higher potential strength and the

concomitance pozzolanic reactions of the HVFA concrete is true. This shift in
195
emphasis is accompanied by increasing permeability of all HVF A concretes at the end

of 504 days even if the concrete is under moist curing. Considerable permeance had
occurred at the end of 504 days, presumably due to the pre-oven dry condition ( at
105°C for 24 hours) may caused some damage to the microstructure. However, the
variation in porosity indicats that higher temperature ( 105°C ) has a deal effect on
the arrangement of these growing hydration products. Furthermore, the maturity is a
function of the composition and properties of the cementitious material, curing
temperature, and of the age of the concrete. The maturity of the concrete at later
ages of the testing also influences the results of the permeability test [9]. This result is
in agreement with the effects of heat curing on cement hydration observed by Goto et
al [94]. They stated that the porosities of samples cured at 60°C were lower than
those cured at 27 °C and were related to higher permeability also in the former.
The relative rates of development of permeability ofHVF A concretes cured in

water and air curing at various ages are shown by the Table 7-10.
Table 7-10 Influence of air curing on permeability

( Unit: % )
Days 28 90 540
Curing Water Air Water Air Water Air

regIme cunng cunng cunng cunng cunng cunng
350F 100 189 100 285 100 417
350SF 100 180 100 154 100 432
350S 100 186 100 126 100 279
450F 100 189 100 279 100 589
450SF 100 187 100 141 100 283
450S 100 185 100 157 100 225
196
It is clear from these data that the permeability of the HVF A concrete was
directly related to the moisture at any give time. All results also indicate that concrete
curing in water is superior to that air drying curing. This advantage has also been
observed for both strength development rate and dynamic modulus of elasticity of
concrete.
The results indicate that the use of air curing of all HVFA concretes exhibited
a marked increase of permeability with continued air drying, even though the
concrete had higher strength. In general, the coefficient of permeability of concrete
incorporating both fly ash and silica fume as partial replacement for cement
performed lower than that concrete with cement alone replaced partially by fly ash.
The percentage increase of permeability coefficient with air cured specimens

is about 86 % of average values for all HVF A concretes when compared with the
moist cured specimens at the age of 28 days. At the age of 90 days, even at the
higher measured strength, considerably higher permeability coefficient was found.
The percentage increase on the coefficient of permeability for dry cured specimens in
HVFA concretes are higher, ranged from 125 % to 489 % for both types of
cementitious material contents when compared with the moist cured concretes at the
age of 18 months. However, the more dense and strong the concrete, the lower is the
value of its increase. The resistance of porous media to fluid is largely a function of
capillary wall friction caused by fluid movement and is therefore affected by pore
diameter, interconnectivity of the pore system, and tortuosity of individual pores
[118]. It has been shown from the pore size distribution in Table 7-8 that air curing
significantly increases macro-pore diameter and hence by implication effective pore
diameter of concrete. Furthermore, drying the specimen increases its permeability,
probably because shrinkage may rupture some of the gel between the capillaries and
thus open new passage to air. The results of this study agree to Dhir and Byars
[118,119,120] in their fly ash concrete investigation that air curing significantly
increases large diameter porosity and hence by implication effective pore diameter of
concrete.
197
In concrete systems, the values of the coefficient of permeability decrease

very substantially with a decrease in the water / cement ratio. Specially, at a water /
cement ratio of 0.75, the coefficient of permeability is typically 10-10 ( m2 ) and this
would be considered to represent concrete with a high permeability. At a water /
cement ratio of 0.45, the coefficient of permeability is typically 10- 11 or 10- 12 ( m2 ),
permeability of an order of magnitude lower than the last value, considered to

represent concrete with a very low permeability [69]. In the present investigation, all
HVFA concretes' coefficient of permeability values were low, ranging from 0.144 x
16 2
10-16 to 0.846 X 10- (m ) for both types of curing conditions at the end of 504 days
measunng.
7-3-2. Permeability and strength development of high volume fly ash

concrete
In Figs 7-27 and 7-28, the coefficients of permeability, Per, are plotted against the
corresponding compressive strength, fe, of the data at the ages of 28 days, 90 days,
180 days, and 18 months for moist and dry cured specimens of both types of
cementitious material content HVF A concretes. Regression analysis of the data
shown in these figures produced the following relationships :
Water curing condition specimens

fe = 23. 71 Per -0.25 ( 28 days) ( r = - 0.93 ) ------- ( 7 - 5 )
fe = 1.43 Per -1.78 ( 90 days) (r = - 0.83 ) ------- (7 - 6 )
tC = 5.48 p er -l. 09 ( 180 days) (r = - 0.96 ) ------- ( 7 - 7 )
fe = 71.14 Per -0.07 (18 months) (r = 0.21 ) --------- ( 7 - 8 )
Air curing condition specimens
fe = 843.6 Per -2.25 (28 days) ( r = - 0.83) ------- ( 7 - 9 )
fe = 7004.4 Per -2.36 (18 months) ( r = 0.86 )--------- ( 7 - 10 )
where fe is the compressive strength in MPa, and Per is the coefficient of permeability
in 10- 16 ( m2 ).
198
80r-----~~------------------------------
70
~
~
28
~ D I
water
" 60
ts:: 90
(\)
~ 50 0
.-
(\)
;;>
1Zl
1Zl
(\)
180
0
~40 540
0 B
U
30
20 L -_ _~_ _ _ __ L_ _ _ _~_ _~L__ _~_ _ _ _~_ _ _ _~~~
o 0.1 0.2
16
0.3 0.4
Permeability Coefficient, 16 (nl )
Figure 7-27 Relationship between permeability and compressive strength of
HVFA concrete-water curing
80
28 540
EI
70
~
~
" 60
ts:: ~-
(\)
~ 50
.-(\)
;;>
1Zl
1Zl
E'~ 40 +++
0
u
30 +
20L---~--L---~--~--~--~--~---L--~--~
o 0.2 0.4 0.6
16 2
0.8 1
Permeability Coefficient, 16 (m )
Figure 7-28 Relationship between permeability and compressive strength of

HVFA concrete-air curing
199
It is obvious from the Fig 7-27, the higher degrees of hydration of HVF A
concretes at the age of 540 days, the greater is the coefficient of permeability. In
view of the positive correlation factor ( r = 0.21 ) found in the equation relating to
the concrete at 18 months with water curing, it is suggested that the correlation
coefficient was very low. It is relieved that the compressive strength is still controlled
by the porosity, the more the meso-pores of the concrete, the stronger is the
compressive strength.
7-3-3. Permeability-porosity relation in high volume fly ash concrete
As already demonstrated, the strength development of concrete is not a simple

function of its porosity, but depends also on the size and distribution of the pores.
Thus, it follows that the permeability of hardened cement concrete is controlled by its
capillary porosity.
The relationship between the coefficient of permeability, Per, and meso-pore

volume, Meso, is generally by the following equation:
Per = a Mesob --------------- ( 7-11 )
16 2
where Per is the coefficient of permeability of HVF A concrete in 10- m , Meso is the
meso-pore volume ofHVF A concrete in %, and a and b are constants.
The relationship between coefficient of permeability and the percentage of

meso-pore for moist and dry cured specimens with both types of cementitious
material content, at ages of 1 day to 18 months, is shown in Fig 7-29. Regression
analysis of the data shown in the figure produced the following relationship :
Per = 35.74 Meso-I.23 (Moist + Dry) (r = - 0.85 ) ---- ( 7-12 )
As shown in figure, there is a reasonable inverse correlation between the

coefficient of permeability, Per, and meso-pores, Meso, of HVF A concrete which is
curvilinear, in other words, the permeability coefficient of concrete decreases with an
increase in the meso-pore volume. The relationship is observed to be influenced by
200
Chapter 7 Test Results and Discussion of Part II ]01
the air cured specimens. Although these correlation coefficients are somewhat low,
there is a significant correlation at the 99 % confidence level between the coefficient
of permeability and the meso-pores of HVFA concretes with both types of
cementitious material contents under both curing conditions. Now, satisfactory
explanation is apparent for this, however it is likely to be connected to the
permeability of the hardened cement concrete when an adequate moist, and the
degree of hydration, and hence lower coefficient of permeability was obtained.
1.6 ,-----------------------~
350F 350SF 350S 450F 450SF 450S

~1.4 x <) o + •
~
'D
'-" 1• 2
....
'0
.......
1
5u
.....
IE 0.8
Q.)
o
U
.,e. 0.6
.....
.....
..... +
~
§ 0.4 +
Q.)
~
0.2 .
•
o L-_~_~ _ _ ~_ _L_~_ _~_~_~_ _~_~
o W ~ W W 100
Meso-pore, %
Figure 7-29 Relationship between permeability and meso-pores ofHVFA
concrete
7-4. WATER ABSORPTION OF HVFA CONCRETE
Absorption is an important concrete characteristic because it strongly influences the

chemical stability, hardness, strength, and thermal properties of the concrete. Since
the hardened cement concretes always contain more or less pores inside, practically
all dry concretes are capable of absorbing water. The amount of the absorbed water
depends primarily on the abundance and continuity of the pores in the concrete,
whereas the rate of absorption depends on the size and continuity of these pores.
201
[121]. Other important factors, such as the curing history, water / cement ratio,
aggregate characteristics, air content, cement type and fineness, and surface
carbonation may also an effect on absorption.
Absorption is usually measured by drying a specimen to a constant mass,

immersing it in water, and measuring the increase in mass as a percentage of dry
mass. In the present study, an absorption test on whole 75 mm x 100 mm core
specimens as prescribed by BS 1881 : Part 122 : 1983, drying at 105°C for 72 hours
and immersion in water for 30 minutes, are determined. Each value represents the
mean value for three specimens. All the specimens and performance procedure have
been described in Chapter six. The test specimens were subj ected to different curing
and tested at the age of 28 days, 90 days, 180 days and 18 months.
In this section a detailed analysis of the water absorption on HVF A concretes

will be discussed, and on its result will be described about the effect of different
replacement method and curing regime on the characteristics of absorption. A
relationship between the permeability of air and the percentage of water absorption
for concrete was described.
7-4-1. Effect of curing condition and different replacement methods
Under water curing for various periods, the results of the water absorption of HVF A
concretes with both types of cementitious material contents are shown in Table 7-11
and plotted in Figs 7-30 and 7-3l. When compared at equal 28 day and 90 day
strength for the wet-cured specimens the concretes made with 350SF, 450SF had
greater water absorption, whereas for the Mix-350S and Mix-450S, with fly ash both
as cement and sand replacement, had the lower water absorption. At the age of 180
days, the water absorption of all the HVF A concrete was reduced so that a lower
valued for both types of cementitious contents was measured. A more noticeable
reduction for concrete 350SF and 450SF, relative to the corresponding 350F and
450F, was observed.
202
Table 7-11 High volume fly ash concrete water absorption ( Unit: % )
Curimg Water curing Air curing
Days 28 90 180 540 28 180 540
350F 1.801 1.661 1.120 1.219 2.190 1.978 1.768
350SF 1.844 1.753 1.078 1.058 2.269 1.244 1.873
350S 1.020 0.963 0.807 0.977 1.317 1.240 1.809
450F 1.578 1.499 1.024 1.053 1.889 1.378 1.810
450SF 1.631 1.534 0.943 1.072 1.672 1.347 1.994
450S 1.015 0.961 0.784 1.082 1.293 1.149 1.809
For the long term wet curing, the water absorption of all HVF A concretes
have appeared similar behaviour to that air permeability. The concretes exhibited
slight increments in water absorption at the age of 18 months when compared to the
age of six months with the same conditions. However, the higher potential strength
and the consequent pozzolanic reaction of the concretes do exhibit good ability
characteristics to resist water penetration.
In moist concretes, pores of different size are always occupied so that the
absorption characteristics of HVF A concrete under water curing as well as drying
condition varied depending on its pore volume and pore size distribution. From the
preceding discussion on porosity, the meso-pore volumes of the concrete made of
350S are more than those of the concrete of 350SF and 350F. The result may
reasonably suggest therefore that water does not easily move through the very small
gel pores and that absorption is controlled by an interconnecting network of capillary
pores [13]. Furthermore, as mentioned previous section, using fly ash to replace both
cement and sand partially might have a combined effect of both separate
replacements.
203
Chapter 7 Test Results and Discussion o{Part II 204
3 r-----------------------------------------
350F 350SF 3508
/ 0 0
2.5
0.5L-J.----~--.L--'---'----'-....L.----------L---------'----L---'-.l.......J
2S 90 ISO 540
Age (Days)
Figure 7-30 The water absorption ofHVFA concrete-water curing

( 350 kglm 3 mixes )
3 ~--------------------------------------------
450F 4508F 4508

•
2.5
';:!(
0
§~
..... 2
e-o
rIl
~.... -+-==--~-
-----========= ~,--
1.5
2
~
1 •
0.5 L2...i.S----L-----'------'---'--9LO----l-iS-0------'----'-----'-5..i--J
40
Age (Days)
Figure 7-31 The water absorption of HVFA concrete-water curing

( 450 kg/m 3 mixes)
204
Another important consequence of the concretes' 450 kg/m3 mixes, at the age
of 18 months, was that the difference in water absorption was very small and
consistent trends could be identified.
The relative rates of development of the water absorption of HVF A concretes

cured in the water and air drying at various ages are shown in Table 7-12.
Table 7-12 Influence of air curing on water absorption

( Unit: % )
Days 28 180 540

Curing Water Air Water Air Water Air
regime curing curing curing curing curing curing
350F 100 122 100 145 100 177
350SF 100 123 100 177 100 115
350S 100 129 100 185 100 154
450F 100 120 100 172 100 135
450SF 100 103 100 186 100 143
450S 100 127 100 167 100 147
The percentage increase of water absorption with air curing specimens is

about 20 % of average values for all HVFA concretes when compared with the moist
cured specimens at the age of 28 days. However, the percentage increased of water
absorption increases with age of all mixtures in air curing specimens. The percentage
increase of water absorption for all dry cured specimens ranged from 15 % to 77 %
for both types of binder contents when compared with the moist cured specimens at
the age of 18 months.
From the results, the curing regime had a much more significant effect on the
water absorption than did the cementitious material, the concretes cured under dry
conditions having a considerably greater water absorption [122]. These increases in
the water absorption could be due to the change in microhardness and retrogression
of the concretes due to air curing regime. For the most part, concrete after mixing
205
with water becomes a solid / fluid suspension and flowability gradually disappears as
free water in concrete is gradually reduced. At the same time, the rigidity of the
concrete' becomes stronger and begins to develop its strength. The reduction of free
water is to caused by the formation of hydrates such as ettringite and C-S-H which
need a large amount of water. The inner structure, due to the formation of hydrates,
becomes a network which contributes to strength [13]. However, the pre-
conditioning regime by air curing may cause a variety of change on internal parts of
concrete. Changes in mechanical properties and porosity with time have been
discussed previously for these system. The present investigation has suggested that
this change in water absorption may due to an accumulation of dense hydration
products consisting of C-S-H, ettringite, Ca(OH)2, and hydrated aluminates at earlier
periods around the unhydrated cementitious grains during air curing [123]. It is
possible that the interaction of hydration products with fly ash or silica fume blends
partially would cause increased loss of moisture and would be expected to lead to
more extensive drying shrinkage cracking, which would be extended to form
continuous channels around the unhydrated cementitious grains with time, thus
allowing gel to imbibe more water. The increase of water absorption also helps to
explain the lower strength, dynamic modulus of elasticity, and pulse velocity in the air
cured concretes due to the combined effect of a higher permeability and water
absorption.
7-4-2. Absorption-permeability relation of HVFA concrete
The relationship between absorption and the coefficient of permeability for both
curing conditions and cementitious material contents, at the age of 504 days is shown
in Fig 7-32. Regression analysis to establish the relationship between water
absorption and the permeability coefficient from the data shown produced the
following equation :
Wabs = 2.21 Per°.40 (moist + dry) (r = 0.94 ) ---------- ( 7-13 )
206
Irrespective of the curing condition or type of cementitious content, the water

absorption, Wabs, increases approximately with the cube root of its coefficient of
permeability, Per. There is a reasonable curvilinear relationship between these two
parameters, although somewhat scatters occur by air cured specimen. These data
indicate a significant correlation at the 99 % confidence level between water
absorption and air permeability ofHVFA concretes at the later period.
2.5 r - - - - - - - - - - - - - - - - - - - - -_ _~
• 1'.
1
540 540
water air
0.5 oLL-'----L--'----L--------1---'-----'-------'-------'1
0.2 0.4 0.6 0.8
16 2
Penneability Coefficient, 10- (m )
Figure 7-32 Relationship between permeability and water absorption ofHVFA

concrete
207
Chapter 7 Test Results and Discussion orPart II 208
7-5. SOLID PHASE ANALYSIS: X-RAY POWDER

DIFFRACTION
The crystal structure of different minerals always produces a characteristic diffraction

pattern, whether that crystal is present in the pure state or as one constituent of a
mixture of materials. This fact is the basis for the X-ray powder diffraction analysis.
The following section describes the X-ray diffraction to identify the crystalline
phase and also to investigate the crystal structure of some of the phases. Routine
qualitative scans of the various mixtures readily identified the predominant phases,
calcium hydroxide, Ca(OH)2, and calcium silicate hydrate ( C-S-H gel ), together
with evidence of unhydrated concrete minerals such as silicon dioxide ( a-quartz ),
calcite ( CaC03 ) and gypsum ( C- S -H2 ).
The test specimens were subjected to different curing conditions and analysed
at the age of28 days, 90 days, and 18 months. For comparison the hydrated powders
of normal ope and HVFA concretes are also included. These studies were done for
the OPC and HVF A concretes made of 450 kg/m3 mixes. All the samples were mixed
and prepared as has been described in Chapter three and Chapter six
7-5-1. X-ray diffraction of normal OPC and HVFA concretes with

water curing
The hydration products of normal OPC and HVF A concretes with and without silica
fume under water curing with curing time are seen from the X-ray diffraction patterns
in Figs 7-33, 7-34, 7-35 and Tables 7-13, 7-14, 7-15.
In the above figures all the X-ray diffractograms have been plotted with
identical scales for both intensity ( I ) and 28, where 8 is the Bragg's angle. Overall
from the figures, a maximum intensity ( 100 % ) peak corresponding to the a-quartz
appears at 3.351 A for both OPC and HVFA concretes throughout the curing
hydration period.
208
Chapter 7 Test Results and Discussion of Part 1I 209
Unfortunately the C-S-H gel is poorly crystalline and amorphous to X-ray

diffraction. Hence this technique is capable only of identifying the crystalline reaction
products. Apart from the portlandite, Ca(OH)2, a-quartz ( Si02 ), calcite ( CaC0 3 )
and perhaps carbo aluminate phase also appear in all samples independent of the
different replacement methods. The a-quartz is believed to be from the aggregate
used for compounding the concrete while, the calcite and carboaluminate phase might
be as a result of the carbonation of portlandite during curing.
However, for all mixtures, the intensity of the X-ray reflections corresponding
to the unhydrated compounds strongly decreased as a result of the hydration reaction
under water curing regime.
Hydration products in the HVF A concretes were found to be different

compared to that normal OPC concrete. As expected from the Fig 7-33, the concrete
made of 450C with hydrated cement alone showed that the extent of hydration of
calcium hydroxide ( CH) ( 450C, *1, d = 4.928 ~ 1= 10, *21, d = 1.799 A, 1=4,
*22, d = 1.69 ~ 1= 1 ) exceeded those of concrete made ofHVFA. Additionally, the
amount of calcite phase in the OPC concrete is slightly higher than the concrete
containing HVF A.
Up to three months curing further changes of the phase composition of the

mixtures is detected by X-ray diffraction. It can be seen from Fig 7-34 that the
diffraction peak of calcium hydroxide, Ca(OH)2, existing in the OPC concrete (
450C, *1, d = 4.928~ 1=8) and in the HVFA samples ( 450F, *1, d = 4.923 A, 1=
3 ) ( 450SF, *1, d = 4.928 ~ I = 1 ) become low. The peaks of calcium hydroxide,
Ca(OH)2, in Mix-450S completely disappear at 90 days. As shown by the above
results with the addition of mineral admixture is able to hydrate further with time.
When the hydration progresses the peaks of non-reactive phase decrease due to the
activation of fly ash and silica fume with liberating calcium hydroxide, Ca(OH)2.
Therefore, the calcium hydroxide is consumed at all ages of hydration [4].
The long term hydration reaction of X-ray diffraction pattern on HVF A

concrete is shown in Fig 7-35. As expected, the hydration products of calcium
209
hydroxide, Ca(OH)2, completely disappears during 18 months for all HVFA

concretes. Again, the presence of any pozzolanic material tends to consume any
liberated calcium hydroxide as shown from the X-ray diffraction patterns.
210
Table 7-13 X-ray diffraction of ope and HVFA concrete-Water curing

( 28 days of hydration )
Hydrates 450C 450F 450SF 450S

products
CH peakl, 29 = 18.0 peakl, 29 = 17.94 peakl, 29 = 17.94 peakl, 29 = 18.02
d = 4.928, 1= 10 d = 4.944, I = 7 d = 4.944, 1=4 d = 4.923, 1=3
peak8, 29 = 28.66 peak6, 29 = 28.6

d=3.115,I=2 d=3.121,I=1
peakl2, 29 = 34.06 peakl0, 29 = 34.02 peak6, 29 = 34.0 peak7, 29 = 3~.1

d = 2.632, I = 10 d = 2.635, 1=4 d=2.637,I=3 d= 2.629, 1=3
peakl9, 29 = 47.12 peak17, 29 = 47.06 peak13, 29 = 47.0 peak13, 29 = 47.16

d=1.929,I=4 d = 1.931, 1=2 d = 1.933, 1=2 d= 1.927, 1=3
peak21, 29 = 50.74 peakl9, 29 = 50.76 peakl5, 29 = 50.6 peak15, 29 = 50.64

d = 1.799, 1=4 d = 1. 799, I = 2 d = 1.804, I = 2 d = 1.803, 1=2
a-quartz peak2, 29 = 20.82 peak2, 29 = 20.78 peak2, 29 = 20.74 peak2, 29 = 20.82

d = 4.267, I = 16 d = 4.275, 1=23 d = 4.283, 1=27 d = 4.267, 1=31
peak5, 29 = 26.6 peak4, 29 = 26.56 peak4, 29 = 26.56 peak3, 29 = 26.6

d = 3.351, 1=100 d = 3.356, I = 100 d = 3.356, 1= 100 d = 3.351, 1=100
peak13, 29 = 36.52 peakll, 29 = 36.44 peak7, 29 = 36.48 peak8, 29 = 36.52

d = 2.46, I = 10 d = 2.466, I = 17 d = 2.46, 1=6 d = 2.46, 1= 16
peak14, 29 = 39.4 peakl2, 28 = 39.4 peak8, 29 = 39.36 peak9, 29 = 39.4

d = 2.287, I = 6 d = 2.287, I = 6 d = 2.289, 1= 5 d = 2.287, I = ~
peakl7, 29 = 42.42 peakl4, 29 = 42.38 peakll, 29 = 42.38 peakl1, 29 = 42.42

d=2.131,I=5 d = 2.133, I = 8 d = 2.133. 1=5 d = 2.131, 1=13
peak18, 29 = 45.46 peakl6, 29 = 45.72 peak16, 29 = 45.72 peak12, 29 = 45.76

d = 1. 995, I = 4 d=1.984,I=3 d=1.984,I=5 d = 1.983, 1=7
peak20, 29 = 50.08 peakl8, 29 = 50.04 peakl4, 29 = 50.08 peak14, 29 = 50.1

d = 1.821, I = 10 d = 1.823, 1= 14 d= 1.821,1= 18 d = 1.821, I = 25

d = 1.674, 1=6 d = 1.675, I = 5 d = 1.675, 1=4 d = 1.673, 1= 3
peak24, 29 = 59.9 peak22, 29 = 59.88 peak19, 29 = 59.88 peakl7, 29 = 59.92

d = 1.544, I = 8 d = 1.545, I = 8 d = 1.545, 1= 8 d = 1.544, I = 8
CaCD3 peak9, 29 = 29.32 peak7, 29 = 29.32 peak5, 29 = 29.42
d= 3.046, 1=3 d = 3.046, 1=2 d = 3.036, 1=2
211
Table 7-14 X-ray diffraction ofOPC and HVFA concrete-Water curing

( 90days of hydration )
Hydrates 450C 450F 450SF 450S

products
CH peakl, 28 = 18.0 peakl, 28 = 18.02 peakl, 28 = 18.0 ---
d = 4.928, 1= 8 d = 4.923, 1=3 d=4.928, 1= 1
peak6, 28 = 28.62 --- --- ---

d = 3.119, 1= 1
peak8, 28 = 34.08 peak5, 28 = 34.08 peak6, 28 = 34.1 ---

d = 2.631, I = 6 d=2.631,I=3 d = 2.629, 1= 1
peakl5, 28 = 47.08 peak11, 28 = 47.12 --- ---

d=1.93,I=2 d=1.929,I=1
peakl7, 28 = 50.74 peakl7, 28 = 50.72 peak 13 , 28 = 50.6 ---

d=1.799,I=2 d = 1.8, 1=1 d = 1.804, I = 1
a.-quartz peak2, 28 = 20.82 peak2, 28 = 20.84 peak2, 28 = 20.84 peakl, 28 = 20.78

d = 4.267, 1= 12 d = 4.262, I = 16 d = 4.262, 1= 17 d = 4.275, I = 16

d = 3.351, 1= 100 d = 3.349, 1= 100 d = 3.346, 1= 100 d = 3.354, 1= 100

d = 2.46, 1=3 d = 2.459, 1= 5 d = 2.459, 1=7 d=2.467,I=18
peaklO, 28 = 39.44 peak7, 28 = 39.44 peak8, 28 = 39.44 peak5, 28 = 39.4

d = 2.285, I = 3 d = 2.285, I = 3 d = 2.285, I = 5 d=2.287,I=4
peak 13 , 28 = 42.44 peak9, 28 = 42.44 peaklO, 28 = 42.4 peak7, 28 = 42.38

d = 2.13, 1=3 d=2.13,I=4 d = 2.132, I = 4 d = 2.133, I = ~
peakl4, 28 = 45.76 peaklO, 28 = 45.74 peak11, 28 = 45.78 peak8, 28 = 45.72

d = 1.983, 1=2 d = 1. 984, I = 3 d = 1. 982, I = 3 d = 1. 984, I = 2
peakl6, 28 = 50.12 peak12, 28 = 50.12 peak12, 28 = 50.1 peak9, 28 = 50.06

d = 1.82, 1=9 d = 1. 82, I = 11 d = 1.821, 1=9 d = 1. 822, I = 6
peakl9, 28 = 59.92 peak15, 28 = 59.92 peakl7, 28 = 59.96 peak12, 28 = 59.88

d = 1. 544, I = 6 d = 1. 544, I = 8 d = 1.543, 1= 15 d = 1.545. 1=4
212
Table 7-15 X-ray diffraction on HVFA concrete-Water curing

( 18 months of hydration)
Hydrates 450F 450SF 450S

products
CH --- -- ---
a-quartz peak2, 28 = 20.78 peak1, 28 = 20.78 peak1, 28 = 20.76
d = 4.275, 1=22 d = 4.275, 1= 24 d = 4.279, I = 23
peak3, 28 = 26.6 peak3, 28 = 26.58 peak2, 28 = 26.54

d = 3.351, 1=100 d = 3.354, I = 100 d = 3.359, 1= 100
peak8, 28 = 36.52 peak8, 28 = 36.5 peaklO, 28 = 36.42

d=2.46, 1=9 d = 2.462, I = 7 d=2.467,I=5
peak9, 28 = 39.42 peak 10 , 28 = 39.38 peak11, 28 = 39.38

d = 2.286, I = 10 d = 2.288, I = 4 d = 2.288, I = 9
peak11, 28 = 42.42 peak13, 28 = 42.42 peak15, 28 = 42.32

d=2.131,I=5 d = 2.131, 1=4 d = 2.136, I = 5
peak12, 28 = 45.72 peak14, 28 = 45.74 peak16, 29 = 45.72

d = 1.984, 1=4 d = 1.984, 1=4 d = 1. 984, I = 4
peak13, 28 = 50.1 peak15, 28 = 50.04 peak18, 28 = 50.2

d=1.821,I=23 d = 1.823, I = 8 d = 1.823, 1= 18
peak17, 28 = 59.92 peak18, 29 = 59.88 peak20, 28 = 59.82

d = 1.544, I = 6 d = 1.545, 1=7 d = 1.546, 1=5
213
o
•
2000~
~
I
1I
1
I
450C
1000
450F
1000 o, N
x
U
I
o
xU
UJ
::r
U l
d
N
-g ::r
;::l
'j'"
U
d
o, o
/1 "
o ~-LJLJ t i , I I I I I : I . i
2000
II
450SF
N
1000 ::t:
U ~'j'"U
::r
d
0,
Ii
1000
450S o,
o o
" "
o, o
II
50 60 70
10 20 30 40
Degrees 2-Theta
Figure 7-33 X-ray ditTractograms for ope and HVFA concrete-Water curing
( 28 days hydration)
214
~IV;
IV .."
C>
C> C>
C>
C>
'"
o
o
C>
0
IV
C>
C)
C>
C>
o
C>
C> o C> C> C> C>
o 0 0.0 C> C>
C) I I I '. ------.L_~J _ _ L~__'__ ~
](-L- l--------.l~ ,~ I __ ~~_..l- __ --1...- _ _ _ _ L .. \ ---'--
_.
~
(JQ
.t>-
V>
o
.t>-
V>
o
.t>-
V>
0
.t>-
V>
o
91:::1
C/l ~
=
'"'t
~
C/l "T1 "T1 ()
""l
'-l
-..l
I
<> CH <> CH -j ==-=- <> CH ~
~ IV
~
~ C)
:::tl
_<> (x-quartz u-quartz
[~~
<> a-quartz u-quartz
>< ~
ll~
I :::!
'"'t
~
~ ~
::\
-.
Q.
a-quartz (x-quart/. a-quartz u-quanz =c-"-co~~'======~---
s:;.,.
tJ
~ .oO .0 ~ ~C> C;;'
=f"')
..... ~<>
()
:::!
V,
-e
(..>
Q o V,
(JQ C3
::\
\0 ;i a
-,
0 CH CH
~
~o
~
.0
CH
Q. ....,
=
'<
~
~
'"'t 0
(I)
-l~-
~<> a-quartz _0 u-quartz .C> a-quartz cO a-quartz
-.
~
N
...-
VI =-
'<
Q.
0
~
CO
iil
(I)
Ul~
~<> a-quartz .0 a-quanz _0 a-quanz a-quartz
(j ",0 .0
a-quartz
_. =
;i
.....
Q =
I
--l
::r
(I)
iii
~<>
a-quartz
a-quartz °
a-quartz
(x-quartz cO
a-quartz
a-quaru a-quaru
-~
Q.
=
== a-quartz
I f <>0° o.-qllanZ Cl-quartz
~~
0
u-quartz
CH -~Q CH
>
f"')
Q
=f"')
'"'t
.....
~
~
U'I
C)
'"
coO U-quartz
(x-quartz
r~ ~
L" 0
o
"-G""'
(x-quan/.
l
r
1! ~
C
0 '-"0
CH
u-quartz
"-G"""'
,
r'r .o - CH
«-qllart/
-ll~~"
"-quan/.
~
=
.....
~
'"'t Ol .::~> (x-qllarv ~r-=-::~ ~, «-qllart/
t ;;. "'---..
(x-qu;lrl/ l
--1~7- <> (L-quanl
to
f"') C)
I'
=
::!. ;0""
=
(JQ ~.::' u-quartz
"-G"'fu ~~j ~
</
(l-quartz
I~
o "C .... <=-~
;;. y' . «-quartz ~ ~';c: a-quartz -1 5.. '=''''''0 Cl-qlU1rtz
""~' a-qllart/
-oJ II
o
Chapter 7 Test Results and Discussion o{Part II 216
Start 10° End 70° Step:.02° Speed:2°/min Time/Step:01 ° Wavelength:1.541838A (Cu)

4000:,------------------------------~--------~--~~~~----------------- __________________
I)
)
3000-1
450F
1000
o,
450SF
1000j 450S
o, "
1:
1
l
10 20 30 40 50 70
Degrees 2-Theta
Figure 7-35 X-ray difTractograms for HVFA concrete-Water curing

( 18 months hydration)
216
7-5-2. Influence of air cured on X-ray diffraction
Under air curing, HVFA concretes with 450 kg/m3 mixes the X-ray diffraction
patterns tested at the end of 18 months are shown in the Fig 7-36 and Table 7-16.
The results show the remarkable influence of the air dried curing on hydration
products which are responsible for mechanical change.
In all concrete mixtures, the intensity of the X-ray reflections corresponding

to the unhydrated compounds strongly increased as a result of the hydration reaction.
At the end of 18 months, in all ofHVFA concretes, although both silica fume and fly
ash are present, appreciable amounts of hydration products are formed and
consequently the strength is lower than wet curing concretes. Meanwhile, a maximum
intensity ( 100 % ) peak corresponding to the a-quartz phase appears at d = 4.271 A
for Mix-450S.
Much of the carboaluminate phase and high intensity peak were recorded by
X-ray diffraction analysis on all HVF A concretes. CaC03 hydrates noticeably as is
evidenced by the persistent presence of the peak at 42.36 28, 42.38 28, and 42.4 28
in Mix-450F, Mix-450SF, and Mix-450S respectively.
From the comparative analysis of Fig 7-36 marked differences can be seen in
the degree of alteration caused by the carbonation. Although the combination of silica
fume and fly ash formed the silicic acid gels, carbonation results in the formation of
CaC0 3 through the intermediate formation of carboaluminate hydrates and by direct
reaction between calcium hydroxide, Ca(OH)2, and CO2 [103].
217
Chapter 7 Test Results and Discussion of Part II
Table 7-16 X-ray diffraction on HVFA concrete-Air curing

( 18 months of hydration)
Hydrates 450F 450SF 450S

products
CH peakl, 28 = 17.96 peakl, 28 = 17.98 peak3, 28 = 18.02
d = 4.939, 1=6 d = 4.934, I = 4 d = 4.923, 1=3
peak6, 28 = 34 peak6, 28 = 34.02 peak9, 28 = 34.1

d = 2.637, 1=5 d = 2.635, 1= 6 d = 2.629, 1=4
peak12, 28 = 46.98 peak12, 28 = 47.08 peak17, 28 = 47.1

d = 1. 934, I = 2 d=1.93,I=2 d = 1.929, 1=2
peakl4, 28 = 50.7 peakl4, 28 = 50.68 ---

d = 1. 80 1, I = 2 d = 1.801, 1=2
a-quartz peak2, 28 = 20.78 peak2, 28 = 20.8 peak4, 28 = 20.8
d = 4.275, I = 20 d = 4.271, 1=19 d = 4.271, 1= 100
peak3, 28 = 26.56 peak4, 28 = 26.56 peak6, 28 = 26.56

d = 3.356, 1=100 d = 3.356, 1=100 d = 3.356, 1=99
peak7, 28 = 36.46 peak7, 28 = 36.52 peaklO, 28 = 36.44

d = 2.464, 1= 4 d = 2.46, 1=4 d = 2.466, I = 8
peak8, 28 = 39.4 peak8, 28 = 39.4 peakl1, 28 = 39.4

d = 2.287, I = 7 d=2.287,I=6 d = 2.287, 1= 11
peaklO, 28 = 42.36 peak 10 , 28 = 42.38 peak 13 , 28 = 42.4

d = 2.134, I = 6 d = 2.133, I = 5 d = 2.132, 1=43
peakl1, 28 = 45.68 peakll, 28 = 45.78 peakl6, 28 = 45.74

d=1.986,I=4 d = 1. 982, I = 3 d = 1. 984, I = 6
peak 13 , 28 = 50.06 peak13, 28 = 50.06 peak19, 28 = 50.08

d = 1.822, 1=10 d = 1.822, I = 10 d= 1.821, 1= 19
peak17. 28 = 59.84 peak17, 28 = 59.9 peak22, 28 = 59.88

d = 1.546, 1=5 d = 1.544, 1=5 d = 1.545, I = 8
CaC03 peak4, 28 = 29.4 peak5, 28 = 29.38 peak7, 28 = 29.32
d = 3.038, 1=5 d = 3.04, 1=5 d= 3.046, 1=8
peaklO, 28 = 42.36 peaklO, 28 = 42.38 peak13, 28 = 42,4

d = 2.134, 1=6 d = 2.133, 1= 5 d = 2.132, 1=43
218
4000 Start: 10° End:70° Step:.02° Speed:2°'min Time/Step:.01° Wavelength:1.541838A (Cu)
o,
3000-,
J
J
!
2000~
1000~ 450SF
iI o, o
o
u
u
Ll
Ll :r:
u ~
1'" :r: o
d U
'"
o I
o
Il
450S o
u
u'" N
400 t::
:r:
u
'" oIt
.
o
o,.
o
I
o, 0,
oI I I I I I I I I I 1 I r I 1 1 I I I r I I I I I I I
10 20 30 40 50 60 70
Oegrees 2-Theta
Figure 7-36 X-ray ditTractograms for HVFA concrete-Air curing

( 18 months hydration)
219
7-6. CONCLUSION
From the results of the microstructure analyses on HVF A concretes, the following
conclusions on the pore size distribution and hydrated products on OPC and HVF A
concrete, and the mechanisms controlling the permeability and water absorption are
presented.
1. The curing regime has shown a big influence on the microstructure development
ofHVFA concrete.
2. There exists an approximate indirect relationship between strength and pore size
of hardened concretes.
3. Reduction in the volume of large pores was observed with the progress of the
pozzolanic reaction. Higher concrete strengths were generally associated with
lower volume of large pores in the concrete.
4. The addition of silica fume to the mix decreased the accessible pore volumes for
HVF A concrete in comparison with the concrete without silica fume.
5. According to the air permeability and water absorption test, water-cured

specimens are superior to the air-cured specimen. The coefficient of permeability
of concrete incorporating high volume fly ash is very low. In any case, concretes
made with 450 kg/m3 cementitious material content gave impermeabilities, and
hence water absorption, which are much lower than those obtained from
concretes with 350 kg/m3 cementitious content.
6. The higher potential strength of mature concrete are very sensitive to curing
temperature ( pre-oven dry, at 105 0 C, 24 hours) and degree of hydration.
7. The higher permeability and water absorption of air-cured HVF A concretes are
explained by a poor dispersion of the precipitated hydration products during the
hydration process.
220
Chapter 7 Test Results and Discussion o(Part 11 ]21
8. The pronounced influence of the type of hydration products on the pore

structure and on the strength of hardened concrete is caused by differences in the
crystal structure.
9. A considerable amount of the non-reactive phase of a-quartz is due to the

aggregate used in making the concrete.
10. Decrease in the levels of calcium hydroxide phase were seen with advancing
water curing age in all the HVFA concretes, providing evidence for pozzolanic
reactivity of the fly ash. Furthermore, the calcium hydroxide phase is completely
consumed in HVF A at the age of 18 months.
11. Retrogression of HVF A concrete under air cunng IS associated with the
formation of CaC03 from Ca(OH)2.
221
Chapter 8 Conclusions and Recommendations {or Future Research
CHAPTER EIGHT
CONCLUSIONS AND RECOMMENDATIONS

FOR
FUTURE RESEARCH
8-1. OVERALL CONCLUSIONS
The conclusions drawn from this investigation are summarized at the end of each
part, but the overall major conclusions derived from this study are given below.
1. The use of superplasticizer is compatible with HVFA concrete. Portland cement

concrete mixes had a higher demand of HRWR than the HVFA concrete. The
greater the binder content, the higher was the HRWR dosage required to maintain
a constant consistency.
2. With HVF A concrete, increasing the total binder content brings little benefit in
terms of strength development, pulse velocity, and pore size distribution. The
concrete with the lower binder content had, however, a marginally higher pulse
velocity. Using a low binder content is not only economical per se, but also leads
to overcoming the undesirable effects of drying shrinkage.
3. Adequate curing is essential if HVFA concrete is to achieve its full potential.

Further, the differences between specimens cured under wet conditions, and those
exposed to prolonged drying are in general far more significant than differences
between HVF A concretes made with different total binder content. The curing
regime has a significant influence on the engineering properties and
microstructure. With prolonged air drying up to 504 days, HVF A concretes
exhibited a minimum loss of dynamic modulus and pulse velocity.
4. Although the early strength of concrete made with high volume fly ash was low,
all HVF A concretes developed compressive strength at 28 days in the range 30 to
222
Chapter 8 Conclusions and Recommendations tor Future Research
40 MPa, and 40 to 65 MPa at 90 days. The use of fly ash to replace both cement
and sand simultaneously resulted in considerably higher strength values due to the
decreased w / b ratio, and reduced loss of early strength. The presence of silica
fume changes the strength of HVF A concrete due to the combined effect of
accelerating hydration reactions with microstructural changes.
5. Based on strength development, the preference of using HVF A concrete is ( 1 ),

the use of fly ash to replace both cement and sand simultaneously ; ( 2 ), made
with both fly ash and silica fume ; ( 3 ), made by fly ash alone as cement
replacement. Although low lime fly ash has a low reactivity, a proper design
strategy can achieve economic and early strength requirements. The figure below
shows such a strategy to produce HVFA concrete in order to conserve resources.
Low Heat of Hydration ;

Partial
Low Shrinkage;
Replacement
Good Economics.
of
Cement
t
High Strength ;
Low Porosity; Fly Ash High Early Strength ;
Low Swelling; Low Shrinkage;
Good Economics. Good Economics.
Partial
Fly ash Replacement
with of
Silica Fume Cement and Sand
Figure 8-1 Strategies of using fly ash in HVFA concrete construction
223
Chapter 8 Conclusions and Recommendations (or Future Research 22-1
6. The ettringite formed in the early stages of hydration accelerated swelling. This
suggests that swelling begins at a lower critical degree of hydration. A
considerably lower swelling was observed in the HVF A concrete containing silica
fume, and this resulted in a concomitant decrease in the total accessible pore
volume and porosity.
7. Low swelling of HVF A concrete results in higher drying shrinkage. The latter
6
ranged from 550 x 10-6 to 640 X 10- mlm for all HVFA concretes at 504 days.
8. The dynamic modulus results were generally high, with values of the order of 55
GPa at 90 days for HVFA concrete with silica fume. Due to high pozzolanic
activity, the percentage increment in dynamic modulus from 28 days to one year
of HVFA concrete is about 3 ~ 4 times that of the normal OPC concrete. At the
end of 18 months, the results are higher than the ACI Building Code 318-89
calculated values by about 50 % for the 450 kglm3 mixes.
9. The pulse velocity of HVF A concrete reached stable values by about six months.
Both dynamic modulus and pulse velocity give good indication of the extent of
deterioration of concrete affected by prolonged air drying. The average
percentage loss of dynamic modulus or pulse velocity for all HVF A concretes
was very low, about 3 0/0, when compared to the same concretes air cured for 28
days.
10. Concretes stored in air showed higher carbonation depths than those stored in
water. Accompanying carbonation, HVFA concrete showed decomposition of
hydration products, and as a result, concrete becomes porous.
11. The results from the air permeability and water absorption tests are, to some
extent, related to the results from the pore size distribution test. Both these tests
thus seem suitable for predicting changes in the pore size of concrete systems.
The HVF A concrete with wet curing showed low water absorption and
16 16 2
permeability coefficient. The latter ranged from 0.117 x 10- to 0.305 x 10- m
16 16 2
for HVFA concrete cured for 28 days, and from 0.102 x 10- to 0.151 X 10- m
for HVF A concrete cured for 90 days.
Chapter 8 Conclusions and Recommendations {or Future Research 225
12. The results show that porosity is a major factor in controlling the engineering
behaviour of HVF A concrete. There exists an approximative indirect relationship
between strength and pore size of hardened concretes.
13. X-ray diffraction studies made it possible to gain a better understanding of the
microstructural changes induced by these procedures. The formation of
predominant ettringite / gypsum leads to swelling / drying shrinkage. It was
shown that an increase in curing leads to a change in the microstructure of C-S-H
hydrates. Decrease in the levels of calcium hydroxide ( CH ) was seen with
advancing curing age in all the HVFA concretes, providing evidence for
pozzolanic reactivity of the fly ashes.
8-2. RECOMMENDATIONS FOR FUTURE WORK
During this study several interesting points have come up, which could not be
investigated because of the limitations of time. For further work in this direction, the
author makes the following suggestions.
1. An examination of potential accelerators to mcrease early-age strength

development is required.
2. Further investigation of other fly ashes in HVF A concrete is necessary to

determine the optimum parameters in terms of mix proportions, admixture type
and dosage.
3. To extend the use of HVFA in commercial applications the influence on their

properties of extreme curing conditions such as high or low temperature and low
moisture availability should be investigated.
4. Studies oflong-term creep under different curing conditions will be very useful.
5. The use offly ash to replace both cement and sand has improved early strength. It
is necessary to investigate the effect of different exposure regimes such as,
225
Chapter 8 Conclusions and Recommendations {or Future Research 226
sodium chloride solution and sulphate solution, and evaluate the Iong-tenn
durability characteristics ofHVF A concrete.
226
References ]]7
References
1. Bland, C. H. and Sharp, 1. H., "A Conduction Calorimetric Study of Gasifier

Slag-Portland Cement Blends", Cement Concrete Research, 1991, 21, pp 359-
367.
2. Bland, C. H., Sharp, 1. H. and Barker, A. P., "Electron Probe Microanalysis of

Gasifier Slag Cement Pastes", Advances in Cement Research, 1991 /92, 4, No.
16, Oct., pp. 159-166.
3. Bland, C. H., Sharp, 1. H. and Osborne, G. 1., "Small-scale Durability Study of

Gasifier Slag-Portland Cement Blends", Advances in Cement Research, 1994, 6,
No. 21, Jan., pp. 37-43.
4. Bland, C. H. and Sharp, 1. H., "The Chemistry of Portland Cement-Gasifier Slag

Interactions", Advances in Cement Research, 1990, 3, No. 11, July., pp. 91-98.
5. Papadakis, V. G., Fardis, M. N. and Vayenas, C. G., "Hydration and

Carbonation of Pozzolanic Cements", ACI Materials Journal Vol. 89, No.2
March-April 1992, pp. 119-130.
6. Swamy, R. N., "Fly Ash Concrete Potential without Misuse", Materials and
Structures, No. 23, 1990, pp. 397-411.
7. Malthotra, V. M., "Superplasticizer Fly Ash Concrete for Structural

Applications", ACI Concrete International, Vol. 8, No. 12, December 1986, pp.
28-31.
8. Proceedings American Concrete Institute, Index, Vols. 26-45, 1929-1949,
Synopsis, pp. 34.
9. Berry, E. E., Malhotra, V. M. and Carette, G. G., "Investigation of High Volume

Fly Ash Concrete Systems", Prepared by Radian Canada Inc. Mississauga,
Ontario , Canada and Canada Centre for Mineral and Energy Technology (
CANMET ), Ottawa, Ontario, Canada, Final Report, October 1993.
227
References 228
10. Swamy, R. N., "Fly Ash Utilisation in Concrete Construction", Second

International Conference on Ash Technology and Marketing, Barbican Centre,
London, September 16th-21st 1984, pp. 359-367.
11. Kokubu, M., "Fly Ash and Fly Ash Cement", Proceeding of the Fifth
International Symposium of the Chemistry of Cement, Tokyo, 1968, The
Cement Association of Japan Part IV, pp. 75-102.
12. Mehta, P. K., "Pozzolanic and Cementitious By-products as Mineral Admixture

for Concrete - A Critical Review", First International Conference on The Use of
Fly Ash, Silica Fume, Slag and Other Mineral By- Products in Concrete, July 3 1-
August 5 1983, pp. 1- 46.
13. Mindess, S. and Young, J. F., "Concrete", Prentice-Hall, Inc. Englewood Cliffs,
New Jersey, 1981.
14. Malhotra, V. M., "High-Performance Structural Concrete Incorporating Large

Percentages of Fly Ash", Invited Paper for Presentation at the Fifth International
Colloquium on Concrete in Developing Countries, and for publication in the
proceedings, January 3-6, 1994, Cairo, Egypt.
15. Jahren, P., "Use of Silica Fume in Concrete", First International Conference on
The Use of Fly Ash, Silica Fume, Slag and Other Mineral By - Products in
Concrete, July 31 - August 5 1983, pp. 625-642.
16. Lamond, J. F., "Twenty Five Years Experience Using Fly Ash in Concrete", First
International Conference on The Use of Fly Ash, Silica Fume, Slag and Other
Mineral By - Products in Concrete, Canada, July 3 1 - August 5 1983, pp. 47-69.
17. Guillaume, L., "Pozzolanic Activity of Fly Ash in Portland Cement and Slag
Cement", Silicate Inds., Vol. 28, No.6, 1963, pp. 297-300.
18. Bache, H. H., Idom, G. M., Nepper-Christensen, P. and Nielson, J.,

"Morphology of Calcium Hydroxide in Cement Paste", Symposium on Structure
of Portland Cement Paste and Concrete, 1966, pp. 154-174.
228
References 229
19. Kokubu, M. and Yamada, 1., "Fly Ash Cements", The VI International Congress
on The Chemistry of Cement, Moscow, September 1974, pp. 1-47.
20. Clifton,1. R., Brown, P. W. and Geoffery Frohusdorff, "Reactivity of Fly Ashes
with Cement", Cement Research Progress, 1977, pp. 321-341.
21. Swamy, R. N., Sami, A. R. Ali. and Theodorakopoulos, D.D., "Early Strength
Fly Ash Concrete for Structural Applications", ACI Journal, September -
October 1983, pp. 414-4213.
22. Tikalsky, P. 1., Carrasquillo, P. M. and Carrasquillo, R. L., "Strength and

Durability Considerations Affecting Mix Proportioning of Concrete Containing
Fly Ash", ACI Materials Journal, November - December 1988, pp. 501-511.
23. Olek, 1. and Diamond, S., "Proportioning of Constant Paste Composition Fly
Ash Concrete Mixes", ACI Materials Journal, March - April 1989, pp. 159-166.
24. Berry, E. E. and Malhotra, V. M., "Fly Ash for Use in Concrete - A Critical
Review", ACI Journal, March - April 1980, pp. 59-73.
25. Hopkins, C. 1. and Cabrera, 1. G., "The Influence of Pulverised Fuel Ash on the
Workability of Concrete", Second International Conference on Ash Technology
and Marketing, Barbican Centre, London, September 16th - 24th 1984, pp. 393-
397.
26. Sturrup, V. R., Hooton, R. R. and Clendenning, T. G., "Durability of Fly Ash
Concrete", First International Conference on The Use of Fly Ash, Silica Fume,
Slag and Other Mineral By - Products in Concrete, July 31 - August 5 1983, pp.
71-86.
27. Langley, W. S., Carette, G. G. and Malhotra, V. M., "Strength Development and
Temperature Rise in Large Concrete Blocks Containing High Volumes of Low -
Calcium (ASTM Class F) Fly Ash", ACI Materials Journal, July - August 1992,
pp. 362-368.
229
References 230
28. Stodola, P. R., "Performance of Fly Ash in Hardened Concrete", Concrete

International, December 1983, pp. 64-65.
29. Cripwell,1. B., Brooks, 1. 1. and Wainwright, P. 1., "Time Dependent Properties
of Concrete Containing Pulverised Fuel Ash and a Superplasticizer", Second
London, September 16th - 21st 1984, pp. 313-320.
30. Thomas Michael, D. A and Matthews, 1. D., "Performance of Fly Ash Concrete
in UK Structures", ACI Material Journal, Vol. 90, No.6, Nov. - Dec. 1993, pp.
586-593.
31. Mohammed Maslehuddin, Abdulaziz I. AI-Mana, Mohammed Shamim, and

Huseyin Saricimen, "Effect of Sand Replacement on the Early-Age Strength
Gain and long-term Corrosion - Resisting Characteristics of Fly Ash Concrete",
ACI Materials Journal, January - February 1989, pp. 58-62.
32. Bamforth, P. B., "In Situ Measurement of the Effect of Partial Portland Cement
Replacement Using Either Fly Ash or Ground Granulated Blast Furnace Slag on
the Performance of Mass Concrete", Proc. lnstn. Civil Engineers, Vol. 69" Part
2, Sept. 1980, pp. 777-800.
33. Ross, A. D., "Some Problems in Concrete Construction", Magazine Concrete

Research, Vol. 12, No. 38, 1960, pp. 27-34.
34. Alexander, K. M., Wardlaw, 1. and lvanusec, 1, "A 4 : 1 Range in Concrete

Creep when Cement S03 Content, Curing Temperature and Fly ash Content are
varied", Cement Concrete Research, Vol. 16, No.2, Mar. - Apr. 1986, pp. 173-
180.
35. Tikalsky, P. 1. and Carrasquillo, R. L., "Influence of Fly Ash on the Sulphate
Resistance of Concrete", ACI Materials Journal, Vol. 89, No.1, Jan. - Feb.
1992, pp. 69-75.
230
References 231
36. Rezansoff, T. and Stott, D., "Durability of Concrete Containing Chloride Based
Accelerating Admixtures", Canadian Journal of Civil Engineering, Vol. 17, 1990
pp. 102-111.
37. Manmohan, D. and Mehta, P. K., "Influence of Pozzolanic, Slag, and Chemical
Admixtures on Pore Size Distribution and Permeability of Hardened Cement
Paste", Cement, Concrete, and Aggregates, Vol. 3, No.1, 1981, pp. 63-67.
38. Mehta, P. K., "Effect of Fly Ash Composition on Sulphate Resistance of

Cement", ACI Journal, Proceedings Vol. 83, No.6, November - December
1986,pp.994-1000.
39. Tikalsky, P. 1. and Carrasquillo, R. L., "Fly Ash Evaluation and Selection for use
in Sulphate Resistance Concrete", ACI Materials Journal, Vol. 90, No.6, Nov. -
Dec. 1993, pp. 545-551.
40. C Plowman, CEGB Harrogate, "The Chemistry ofPFA in Concrete - A Review

of Current Knowledge", Second International Conference on Ash Technology
and Marketing, Barbican Centre, London, Sept. 16th - 21st 1984, pp. 437-443.
41. Swamy, R. N. and Al-Asali, M. M., "Engineering Properties of Concrete

Affected by Alkali Silica Reaction", ACI Materials Journal, September - October
1988, pp. 367-375.
42. Swamy, R. N. and Al-Asali, M. M., "Expansion of Concrete Due to Alkali-Silica

Reaction", ACI Materials Journal, January - February 1988, pp. 33-40.
43. Georges, C., Alain B., Chevrier, R. L. and Malhotra, V. M., "Mechanical
Properties of Concrete Incorporating High Volumes of Fly Ash from Sources in
the US", ACI Material Journal, Vol. 90, No.6, Nov. - Dec. 1993, pp. 535-544.
44. Bilodeau, A. and Malhotra, V. M., "High-Performance Concrete Incorporating

Large Volumes of ASTM Class F Fly Ash", Published in the Second Edition of
"Advances in Concrete Technology", issued in 1994 by CANMT, Ottawa,
Canada, pp. 508-523.
231
References 232
45. Alasali M. M. and Malhotra V, M., "Role of Concrete Incorporating High

Volumes of Fly Ash in Controlling Expansion due to Alkali-Aggregate
Reaction", ACI Materials Journal, Vol. 88 No.2 March - April 1991, pp. 159-
163.
46. Buttler, F. G., Deeter, M. H. and Smith, G. R., "Studies on the Desiccation and
Carbonation of Systems Containing Portland Cement and Fly Ash" , First
Mineral By-Products in Concrete, Canada July 31 - August 5 1983, pp. 367-380.
47. Browne, R. D., "Ash Concrete - Its Engineering Performance", Second

London, September 16th - 21st 1984, pp. 295-301.
48. Haque, M. N. and Kawamura, M., "Carbonation and Chloride - Induced

Corrosion of Reinforcement in Fly Ash Concrete", ACI, Materials Journal, Vol.
89, No.1, Jan. - Feb. 1992, pp. 41-48.
49. Swamy, R. N. and Tanikawa, S., "Control of Steel Corrosion in Chloride

Contaminated Concrete Through Aron Wall Surface Coating", Durability of
Concrete, Second International Conference, Montreal, ACI SP-126, Canada,
1991, pp. 371-392.
50. Dhir, R. K., Jones, M. R. and Seneviratne, A. M. G., "Diffusion of Chloride into
Concrete, Influence of PF A Quality", Cement and Concrete Research, Vol. 21,
No.6, November 1991, pp. 1092-1102.
51. Haque, M. N., Kayyali, O. A. and Gopalan, M. K., "Fly Ash Reduces Harmful
Chloride Ions in Concrete", ACI Materials Journal, Vol. 89, No.3, May - June
1992, pp. 238-241.
52. Mehta, P. K. and Gj<l>rv, O. E., "Properties of Portland Cement Concrete

Containing Fly Ash and Condensed Silica Fume", Cement and Concrete
Research, Vol. 12,1982, pp. 587-595.
232
References 233
53. Grutzeck, M. W., Scott, A. and Roy, D. M., "Mechanism of Hydration of

Condensed Silica Fume in Calcium Hydroxide Solutions", First International
Conference on The Use of Fly Ash, Silica Fume, Slag and Other Mineral By -
Products in Concrete, Canada, July 31 - August 5 1983, pp. 643-652.
54. Xiaofeng, C., Shanglong, G., David, D. and Steven, L. M., "Role of Silica Fume
in Compressive Strength of Cement Paste, Mortar, and Concrete", ACI
Materials Journal, Vol. 89, No.4, July - August 1992, pp. 375-387.
55. Gj<j>rv, O. E., "Durability of Concrete Containing Condensed Silica Fume", First
Mineral By- Products in Concrete, Canada, July 31-August 5 1983, pp. 695-709.
56. Sellevold, E. J. and Radjy, F. F., "Condensed Silica Fume in Concrete; Water
Demand and Strength Development", First International Conference on The Use
of Fly Ash, Silica Fume, Slag and Other Mineral By - Products in
Concrete,Canada, July 31 - August 5 1983, pp. 677-694.
57. Collins, T. M., "Proportioning High-Strength Concrete to Control Creep and

Shrinkage", ACI Materials Journal, Vol. 86, No.6, November - December 1989,
pp. 576-577.
58. Swamy, R. N., "Design for Durability and Strength Through the Use of Fly Ash
and Slag in Concrete", unplished, March 1997.
59. Swamy, R. N. and Mahmund, H. B., "Mix Proportions and Strength

characteristics of Concrete Containing 50% Low-Calcium Fly ash", Proc. 2nd
International Conference Fly Ash, Slag, Silica Fume and Natural Pozzolnas in
Concrete, 1986, pp. 413-432.
60. Langley, W. S., Carette, G. G. and Malhotra V. M., "Structural Concrete

Incorporating High Volumes of ASTM Class F Fly Ash", ACI Material Journals,
Vol. 86. No.5 September - October 1989, pp. 507-514.
233
References 234
61. Giaccio, G. M. and Malhotra, V. M., "Concrete Incorporating High Volumes of

ASTM Class fFly Ash", Cement, Concrete and Aggregates, CCAGDP, Vol. 10,
No.2 Winter, 1988, pp. 88-95.
62. Aimin, Xu. and Sarkar, S. L., "Microstructual Development in High Volume Fly
Ash Cement System", Journal of Material in Civil Engineering, Vol. 6, NO.1,
February 1994, pp. 117-137.
63. Sivasundaram, Carette, G. G. and Malhotra, V. M., "Mechanical Properties,

Creep, and Resistance to Diffusion of Chloride Ions of Concretes Incorporating
High Volumes of ASTM Class F Fly Ash From Seven Different Sources", ACI
Materials Journal, Vol. 88, No.4, July - August 1991, pp. 407-416.
64. "Measurement of Hardened Concrete Carbonation Depth", RILEM Draft

Recommendation CPC-18, Materials and Structures, Research and Testing,
RILEM, Paris. Vol. 17. No. 102, 1984, pp. 437-440.
65. Garette and Malhotra, V. M., "Early-Age Strength Development of Concrete

Incorporating Fly Ash and Condensed Silica Fume", First International
Conference on The Use of Fly Ash, Silica Fume, Slag and Other Mineral By -
Products in Concrete, Canada, July 31 - August 5 1983, pp. 765-784.
66. Goldman, A. and Bentur, A., "Bond Effects in High Strength Silica Fume
Concretes", ACI Material Journal, V. 86, No.5, Sept. - Oct. 1989, pp. 440-447.
67. Cohen, M. D., "Theories of Expansion in Sulfoaluminate-Type Expansive

Cement; Schools of Thoughts", Cement and Concrete Research, Vol. 13, No.6,
Nov. 1983, pp. 809-818.
68. Kesler, C. E., "Control of Shrinkage, Creep and Temperature Volume Changes
Through Variations of Material Properties", State of Art Report, Technical
Committee No. 25 : Creep, Shrinkage, and Temperature Effects, ASCE-IABSE
Joint Committee on Planning and Design of Tall Building, 15 January 1972, pp.
1-11.
234
References 235
69. Neville, A. M., "Properties of Concrete", Fourth Edition, 1995.
70. Bentur, A., "The Pore Structure of Hydrated Cementitious Compounds of

Different Chemical Composition", Journal of The American Ceramic Society,
Vol. 63, July - August 1980, pp. 381-386.
71. Bentur, A., "Effect of Curing Temperature on the Pore Structure of Tricalcium
Silicate Pastes", Journal of Colloid and Interface Science, Vol. 74, No.2, April
1980, pp. 549-560.
72. Haque, M. N., Langan, B. W. and Ward, M. A., "High Fly Ash Concretes", ACI
Journal January - February 1984, pp. 54-59.
73. Detwiler, R. J. and Mehta, P. K., "Chemical and Physical Effect of Silica Fume
on the Mechanical Behaviour of Concrete", ACI Material Journal, Vol. 86, No.
6, November - December 1989, pp. 609-614.
74. Baalbaki, W., Aitcin, P. C. and Ballivy, G., "On Predicting Modulus of
Elasticity in High Strength Concrete", ACI Material Journal, Vol. 89, No.5,
Sept.- Oct. 1992, pp. 517-520.
75. Kameswara Rao, Swamy, R. N. and Mangat, P. S., "Mechanical Behaviour of

Concrete as a Composite Material", Materiaux Et Constructions, Vol. 7, No. 40
pp. 265-271.
76. Hobbs, D. W., "The Dependence of the Bulk Modulus, Young's Modulus,
Creep, Shrinkage and Thermal Expansion of Concrete upon Aggregate Volume
Concentration", Materiaux Et Constructions, Vol. 4, No. 20, 1971, pp. 107-
114.
77. Lobo, C. and Cohen, M. D., "Effect of Silica Fume on Expansion Characteristics
of Expansive Cement Pastes", ACI Material Journal, Vol. 89, No.5, Sept.- Oct.
1992, pp. 481-490.
235
References 236
78. Popovics, S., Rose, J. L. and Popovics, J. S., "The Behaviour of Ultrasonic
Pulses in Concrete", Cement and Concrete Research, Vol. 20, 1990, pp. 259-
270.
79. Litvan, G. G. and Meyer, A., "Carbonation of Granulated Blast Furnace Slag
Cement Concrete During Twenty Years of Field Exposure", Fly Ash, Silica
Fume, Slag, and Natural Pozzolans in Concrete, SP-91, American Concrete
Institute, Detroit, 1986, pp. 1445-1462.
80. Kokubu, M. and Nagataki, S., "Carbonation of Concrete with Fly Ash and
Corrosion of Reinforcement in 20 Years Tests", Fly Ash, Silica Fume, Slag and
Natural Pozzolans in Concrete-Proceeding of the Third International
Conference, SP-1l4, Vol. 1, American Concrete Institute, Detroit, 1989, pp
315-329.
81. Ewerstson, C. and Prtersson, P. E., "The Influence of Curing Conditions on The
Permeability and Durability of Concrete Results from a Field Exposure Test",
Cement and Concrete Research, Vol. 23, 1993, pp. 683-692.
82. Malami, C., Batis, G., Noulooumbi, H. and Kaloidas, V., "Influence of
Pozzolanic and Hydraulic Cement Additions on Carbonation and Corrosion of
Reinforced Mortar Specimens", Proceeding of International Conference held at
the University of Sheffield, 24-28 July 1994, Edited by R. N. Swamy, Vol. 2 pp.
668-682.
83. Nagataki, S., Ohga, H. and Kim, E. M., "Effect of Curing Condition on the
Carbonation of Concrete with Fly Ash and the Corrosion of Reinforcement in
long-term Tests", Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete,
SP-91, American Concrete Institute, Detroit, 1986, pp 1445-1462.
84. Afridi, M. U. K., Chaudhary, Z. U., Ohama, Y., Demura, K. and Iqbal, M. Z.,
"Carbonation Resistance of Powered and Aqueous Polymer-Modified Mortars
and Effects of Carbonation on Their Pore Structures", Proceedings of
236
References 237
International Conference held at the University of Sheffield, 24-28 July 1994,

Edited by R. N. Swamy, Vol. 2 pp. 974-987.
85. Dalgleish, B. 1., Pratt, P. L. and Moss, R. 1., "Preparation Techniques and the
microscopical Examination of the Portland Cement Paste and C3 S", Cement and
Concrete Research, Vol. 10, No.5, 1980, pp. 665-676.
86. Dent Glasser, L. S., Lachowski, E. E., Mohan, K. and Taylor, H. F. W, "A
Multi-Method Study of C3 S hydration", Cement and Concrete Research, Vol. 8,
No.6, 1978, pp. 733-740.
87. Berger, R. L. and McGregor, T. D., "Influence of Admixture on the Morphology

of Calcium Hydroxide Formed During Tricalcium Silicate Hydration", Cement
and Concrete Research, Vol. 2, No.1, 1972, pp. 43-56.
88. Bentur, A., Berger, R. L., Lawrence, F. V., Milestone, N. B., Mindess, S. and
Young, 1. F., "Creep and Drying Shrinkage of Calcium Silicate Pastes III A
Hypothesis of Irreversible Strains", Cement and Concrete Research, Vol. 9, No.
1, 1979, pp. 83-96.
89. Roy, D. M., Gouda, G. R. and Bobrowsky, A., "Very High Strength Cement
Pastes Prepared by Hot - Pressing and Other High - pressure Techniques",
Cement Concrete Research, Vol. 2 No.3, 1972, pp. 349-366.
90. Rossler, M. and Odler, I., "Investigation on the Relationship between Porosity
Structure and Strength of Hydrated Portland Cement Pastes I. Effect of
Porosity", Cement Concrete Research, Vol. 15 No.2, 1985, pp 320-330.
91. Watson, K. L., "A Simple Relationship between the Compressive Strength and
Porosity of Hydrated Portland Cement", Cement Concrete Research, Vol. 11,
1981, pp. 473-476.
92. Feldman, R. F. and Beaudoin, 1. 1., "Microstructure and Strength of Hydrated

Cement", Cement and Concrete Research, Vol. 16, 1976, pp. 389-400.
237
References 238
93. Sersale, R., Cioffi, R., Frigione, G. and Zenone, F., "Relationship between
Gypsum Content, Porosity, and Strength of Cement", Cement and Concrete
Research, Vol. 21, No.1, 1991, pp. 120-126.
94. Goto, S. and Roy, D. M., "The Effect of W / C Ratio and Curing Temperature
on the Permeability of Hardened Cement Paste", Cement and Concrete
Research, Vol. 11, 1981, pp. 575-579.
95. Kung, J. H., "Drying Shrinkage and Microstructure of Hydrated Calcium

Silicate", Ph.D. Thesis, University ofTIlinois, 1978.
96. Young, J. F., "Drying Shrinkage of Plain Concrete", unpublished, 1981.
97. Darmawan, L., Richard, L. B. and Young, J. F., "Simple Method for Measuring
Water Permeability of Concrete", ACI Materials Journal, Vol. 86, No.5,
September - October 1989, pp. 433-439.
98. Lea, F. M., "The Chemistry of Cement and Concrete", 1956, Second Edition.
99. Kantro, D. L., Weise, C. H. and Brunauer, S., "Symposium on Structure of

Portland Cement Paste and Concrete", Highway Research Board Special Report
1966.
100.Diamon, M., Abo-el-enein, S. A., Hosaka, G., Goto, S. and Kondo, R., "Pore
Structure of Calcium Silicate Hydrate in Hydrated Tricalcium Silicate", Journal
of The American Ceramic Society, Vol. 60, Mar. - Apr. 1977, pp. 110-114.
1o1. Winslow, D. N. and Diamond, S., "Specific Surface Of Hardened Portland

Cement Pastes as Determined by Small-Angle X-ray scattering", Journal of The
American Ceramic Society, Vol. 57, NO.5 May 1974, pp. 193-197.
102. Jenning, H. M. and Pratt, P. L., "On the Hydration of Portland Cement",
Proceeding of British Ceramic Society, 1979, pp. 179-193.
238
References 239
103.Hansen, T. C., Radjy, F. and Sellevold, E. 1., "Cement Paste and Concrete" ,
Building Materials Laboratory, The Technical University of Denmark, 1973, pp.
233-268.
104.Lachowski, E. E., Mohan, K., Taylor, H. F. W. and Moore, A. E., "Analytical

Electron Microscopy of Cement Pastes. II. Pastes of Portland Cement and
Clinkers", Journal of the American ceramic Society, vol. 63, 1980, pp. 447-452.
105.Lachowski, E. E., "Trimethylsilylation as a Tool for the Study of Cement

Pastes", Cement and Concrete Research, Vol. 9, No.3, 1979, pp. 337-342.
106.Parrott, L. 1. and Taylor, M. G., "A Development of the Molybdate Complexing

Method For the Analysis of Silicate Mixtures", Cement and Concrete Research,
Vol. 9, No.4, 1979, pp. 483-488.
107.Bentur, A. and Young, 1. F., "Simplified method of Determining the Polysilicate

Content in Cementitious Pastes Using Trimethylsilyl Derivatives", Cement and
Concrete Research, Vol. 11, No.2, 1981, pp. 287-290.
108.ChatteIji, S., "Drying Shrinkage of Cement Paste and Concrete: A Reappraisal

of the Measurement Technique and its Significance", Cement and Concrete
Research, Vol. 6, No.1, 1976, pp. 145-148.
109.Cook, R. A. and Hover, K. C., "Mercury Porosimetry of Cement-Based

Materials and Associated Correction Factors", ACI Materials Journal, Vol. 90,
No.2, 1993, pp. 152-161.
1l0.Shi, D. and Winslow, D. N., "Contact angle and Damage During Mercury
Intrusion into Cement Paste", Cement and Concrete Research, Vol. 15, 1985,
pp. 645-654.
111.Winslow, D. N. and Diamond, S., "A Mercury Porosimetry Study of the

Evolution of Porosity in Portland Cement", Journal of Materials, Vol. 5, No.3,
September 1970, pp. 564-585.
239
References 2-10
112. Cabrera, 1. G. and Lynsdale, C. 1., "A new Gas Permeameter for Measuring the
Permeability of Mortar and Concrete", Magazine of Concrete Research, Vol. 40,
No. 144, September 1988, pp. 177-182.
113. British Standards Institution, "Method for Determination of Water Absorption",

BS 1881 : Part 122 : 1983.
114.Cullity, B. D., "Elements of X - ray Diffraction", Second Edition, 1978.
115.Marsh, B. K., Day, R. L. and Bonner, D. G., "Pore Structure Characteristics

Affecting the Permeability of Cement Paste Containing Fly Ash", Cement and
Concrete Research, Vol. 15, 1985, pp. 1027-1038.
116. Roy, D. M. and Gouda, G. R., "Porosity-Strength Relation Cementitious

Materials with Very High Strengths", Journal of The American Ceramic Society
- Discussion and Notes Vol. 56, No. 10, May 1973, pp. 449-550.
117 .Jose Anagel Lechuga Albala and Sebastla Alegre Rossello, "Concrete
Impermeability Study", Second International Conference on Ash Technology
and Marketing, Barbican Centre, London, September 1984, pp. 387-389.
118.Dhir, R. K. and Byars, E. A., "Pulverised Fuel Ash Concrete: Intrinsic

Permeability", ACI Material Journal, Vol. 90, No.6, Nov. - Dec. 1993, pp. 571-
580.
1I9.Dhir, R. K. and Byars, E. A., "Pulverised Fuel Ash Concrete: Near Surface
Absorption Properties", Magazine of Concrete Research, Vol. 43, No. 157, Dec.
1991, pp. 219-231.
120. Dhir, R. K. and Byars, E. A., "Pulverised Fuel Ash Concrete: Permeation
Properties of Cover to Steel Reinforcement", Cement and Concrete Research,
Vol. 23, No.3, 1993, pp. 554-556.
121.Popovics, S., "Concrete-Making Materials", Hemisphere Publishing

Corporation, First Edition, 1979, pp. 167-192.
240
References 2-11
122.Nixon, P. 1., Hardcastle, 1. and Ben-Bassat M., "The effect of differences in the
composition of Portland Cement on the Properties of Hardened Concrete",
Magazine of Concrete Research, Vol. 42, No. 151, Jun. 1990, pp. 59-66.
123.Feldman, R. F., "Significance of Porosity Measurements on Blended Cement

Performance", First International Conference on The Use of Fly Ash, Silica
Fume, Slag and Other Mineral By - Products in Concrete Canada, July 31 -
August 5 1983, pp. 415-433.
2·B

Properties of High Volume Fly Ash Concrete: Hsien-Hsin Hung

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Properties of High Volume Fly Ash Concrete: Hsien-Hsin Hung

Uploaded by

Copyright:

Available Formats

Properties of High Volume Fly Ash Concrete

Thesis Submitted to the University of Sheffield

for the Degree of Doctor of Philosophy

in the Faculty of Engineering

Department of Mechanical Engineering

Thanks are also extended to Professor J. H. Sharp, Head of Department of

CHAPTER ONE: IN"TRODUCTION .................................................................. 1

CHAPTER TWO : LITERATURE REVIEW ..................................................... 7

CHAPTER THREE: EXPERIMENTAL PROGRAMME ............................... 38

CHAPTER FOUR: TEST RESULTS AND DISCUSSION OF PART 1........... 51

CHAPTER FIVE : MICROSTRUCTURE OF HYDRA TED CEMENT

CHAPTER SIX: EXPERIMENTAL PROGRAMME OF PART 11 ............... 149

CHAPTER SEVEN: TEST RESULTS AND DISCUSSION OF PART 11 ..... 162

CHAPTER EIGHT: CONCLUSIONS AND RECOMMENDATIONS FOR

REFERENCES ....•..........•................................................................................... 227

Table 2-1 Chemical composition (% ) offly ash in various countries ...................... 10

Figure 4-1 Compressive strength ofHVFA concrete-Water curing .......................... 55

A. Cement Chemistry Shorthand Notations and Terminology

I L = Ignition Loss MIP = Mercury Intrusion Porosimetry

C. Pore Classification by IUPAC

Unit (Diameter) A run

1-2. BACKGROUND TO THE PROJECT

1-3. AIMS AND OBJECTIVES OF THE INVESTIGATION

cementitious materials varied from 50 to 80 %, while water and portland cement

Part I : Engineering properties of high volume fly ash concrete

The aims and objectives of this part include:

1. To develop mix proportioning methods for the production of high performance

3. To examine the engineering properties of concrete incorporating high volumes of

5. To assess the effect of pozzolanic activity on the carbonation characteristics of

6. To investigate the relationship between vanous engmeenng properties of

Part II : Microstructure mechanism and hydration reaction of high volume fly

The aims and objectives Part II of the investigation are :

1. To study the relationship between microstructure and engineering properties of

properties of high volume fly ash concrete.

4. To investigate the mineralogical composition of calcium hydroxide, Ca(OH)2,

1-4. Thesis layout

The first chapter presents a general introduction to the background and

The second chapter contains a literature review on properties of concrete

There is a great deal of published literature on the properties of concrete containing

The use of fly ash as a pozzolanic material in combination with portland

Fly Ash is a by-product of the combustion of pulverized coal in thermal

2-2. THE CHEMICAL COMPOSITION AND PHYSICAL

Chemical composition of fly ash

On the basis of optical and scanning electron micrograph observations,

3. Magnetite and hematite occur as opaque spheres in high-iron fly ashes.

Table 2-1 Chemical composition ( % ) of fly ash in various countries [11]

2-3. CLASSIFICATION OF FLY ASH

From the standpoint of conspicuous differences in mineralogical composition and

2-4. HYDRATION AND POZZOLANIC REACTION

2-4-1. Hydration of pure cement compounds

Portland cement is a hydraulic cement made by heating a mixture of limestone and

The cement consists essentially of crystalline compounds ( minerals ) of

The chemistry of principal setting, hardening reaction and micromechanism

A : Reaction involving calcium silicates

The hydration reactions between tr-icalcium silicate, di-calcium silicate and

Heat of Hydration AH = - 114 kJ / mol

Heat of Hydration MI = - 43 kJ / mol

The principal hydration product is a colloidal calcium silicate hydrate gel ( C-

The other product of hydration of the silicate minerals is about 20 to 25

B : Reactions involving aluminate

This calcium sulfoaluminate hydrate is commonly called "ettringite". Ettringite

This second product IS called monosulfoaluminate. The calcium

The over reaction ofC3A can thus be represented by :